app-dev-guide.texi

input texinfo @c -*-texinfo-*- @c %**start of header @setfilename semantic-appdev.info @set TITLE Semantic Application Development Manual @set AUTHOR Eric M. Ludlam, David Ponce, and Richard Y. Kim @settitle @value{TITLE}

@c ************************************************************************* @c @ Header @c *************************************************************************

@c Merge all indexes into a single index for now. @c We can always separate them later into two or more as needed. @syncodeindex vr cp @syncodeindex fn cp @syncodeindex ky cp @syncodeindex pg cp @syncodeindex tp cp

@c @footnotestyle separate @c @paragraphindent 2 @c @@smallbook @c %**end of header

@copying This manual documents Application Development with Semantic.

Copyright @copyright{} 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2007 Eric M. Ludlam Copyright @copyright{} 2001, 2002, 2003, 2004 David Ponce Copyright @copyright{} 2002, 2003 Richard Y. Kim

@quotation Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.1 or any later version published by the Free Software Foundation; with the Invariant Sections being list their titles, with the Front-Cover Texts being list, and with the Back-Cover Texts being list. A copy of the license is included in the section entitled ``GNU Free Documentation License''. @end quotation @end copying

@ifinfo @dircategory Emacs @direntry * Semantic Application Writer's guide: (semantic-appdev). @end direntry @end ifinfo

@iftex @finalout @end iftex

@c @setchapternewpage odd @c @setchapternewpage off

@ifinfo This file documents Application Development with Semantic. @emph{Infrastructure for parser based text analysis in Emacs}

Copyright @copyright{} 1999, 2000, 2001, 2002, 2003, 2004 @value{AUTHOR} @end ifinfo

@titlepage @sp 10 @title @value{TITLE} @author by @value{AUTHOR} @vskip 0pt plus 1 fill Copyright @copyright{} 1999, 2000, 2001, 2002, 2003, 2004 @value{AUTHOR} @page @vskip 0pt plus 1 fill @insertcopying @end titlepage @page

@c MACRO inclusion @include semanticheader.texi

@c ************************************************************************* @c @ Document @c ************************************************************************* @contents

@node top @top @value{TITLE}

A @semantic{} application takes the semantic tags generated by semantic parsers then provides useful services to the user. For a list of such applications, @inforef{the Semantic User's Guide, , semantic-user}.

An application developer needs to know @itemize @bullet @item when to invoke the parser to generate or regenerate the tag lists. @item how to access the tag lists. @item how to access information about each tag. @end itemize

This chapter describes @semantic{} fuctions and concepts an application developer needs to know to perform all of the tasks just listed.

@menu * Semantic Tags:: * Searching Tag Tables:: Searching tag tables. * Tags at Point:: Finding tags at point. * Tag Decoration:: Decorating tags * Tag Sorting:: Reorganizing streams. * Tag Completion:: Completing read functions. * Override Methods:: Language dependent functions covering
conversion to text strings, language dependent
queries and local context information * Parser Features:: Application available parser features. * Semantic Database:: Persistant storage of tags. * Idle Scheduling:: Scheduling jobs in idle time. * Example Programs:: Simple programming examples. * Current Context:: Local context analysis. * App Debugger:: Application Debugger * GNU Free Documentation License:: * Index:: @end menu

@node Semantic Tags @chapter Semantic Tags

@include tags.texi

@node Searching Tag Tables @chapter Searching Tag Tables

These functions take some key, and returns information found in a tag table. Some will return one tag (the first matching item found.) Others will return a list of all items matching a given criterion. Most of these functions work regardless of a buffer being in memory or not.

Any specialty search routine that claims to use a function that is an overload method will need to execute in a buffer of the same mode as the tags being searched.

@menu * Breadth Search:: Searching only one level of tags. * Deep Search:: Deep searches into types or argument lists. * Specialty Search:: Specialty Searches. * Custom Search:: Write custom search routines. @end menu

@node Breadth Search @section Breadth Search

Searching the breadth of a list of tags means that only one level of the tags will be searched. If one of the tags is a datatype with additional members, those members are not searched.

@defun semantic-find-first-tag-by-name name &optional table @anchor{semantic-find-first-tag-by-name} Find the first tag with @var{NAME} in @var{TABLE}. @var{NAME} is a string. @var{TABLE} is a semantic tags table. See @dfn{semantic-something-to-tag-table}. This routine uses @dfn{assoc} to quickly find the first matching entry. @end defun

@defun semantic-find-tags-by-name name &optional table @anchor{semantic-find-tags-by-name} Find all tags with @var{NAME} in @var{TABLE}. @var{NAME} is a string. @var{TABLE} is a tag table. See @dfn{semantic-something-to-tag-table}. @end defun

@defun semantic-find-tags-for-completion prefix &optional table @anchor{semantic-find-tags-for-completion} Find all tags whos name begins with @var{PREFIX} in @var{TABLE}. @var{PREFIX} is a string. @var{TABLE} is a tag table. See @dfn{semantic-something-to-tag-table}. While it would be nice to use @dfn{try-completion} or @dfn{all-completions}, those functions do not return the tags, only a string. Uses @dfn{compare-strings} for fast comparison. @end defun

@defun semantic-find-tags-by-name-regexp regexp &optional table @anchor{semantic-find-tags-by-name-regexp} Find all tags with name matching @var{REGEXP} in @var{TABLE}. @var{REGEXP} is a string containing a regular expression, @var{TABLE} is a tag table. See @dfn{semantic-something-to-tag-table}. Consider using @dfn{semantic-find-tags-for-completion} if you are attempting to do completions. @end defun

@defun semantic-find-tags-by-class class &optional table @anchor{semantic-find-tags-by-class} Find all tags of class @var{CLASS} in @var{TABLE}. @var{CLASS} is a symbol representing the class of the token, such as @code{'variable}, of 'function.. @var{TABLE} is a tag table. See @dfn{semantic-something-to-tag-table}. @end defun

@defun semantic-find-tags-by-type type &optional table @anchor{semantic-find-tags-by-type} Find all tags of with a type @var{TYPE} in @var{TABLE}. @var{TYPE} is a string or tag representing a data type as defined in the language the tags were parsed from, such as ``int'', or perhaps a tag whose name is that of a struct or class. @var{TABLE} is a tag table. See @dfn{semantic-something-to-tag-table}. @end defun

@defun semantic-find-tags-included &optional table @anchor{semantic-find-tags-included} Find all tags in @var{TABLE} that are of the @code{'include} class. @var{TABLE} is a tag table. See @dfn{semantic-something-to-tag-table}. @end defun

@node Deep Search @section Deep Search

@defun semantic-brute-find-first-tag-by-name name streamorbuffer &optional search-parts search-include Find a tag @var{NAME} within @var{STREAMORBUFFER}. @var{NAME} is a string. If @var{SEARCH-PARTS} is non-@code{nil}, search children of tags. If @var{SEARCH-INCLUDE} is non-@code{nil}, search include files.

Use @dfn{semantic-find-first-tag-by-name} instead.

@obsolete{semantic-find-nonterminal-by-name,semantic-brute-find-first-tag-by-name} @end defun

@defun semantic-brute-find-tag-by-property property value streamorbuffer &optional search-parts search-includes Find all tags with @var{PROPERTY} equal to @var{VALUE} in @var{STREAMORBUFFER}. Optional argument @var{SEARCH-PARTS} and @var{SEARCH-INCLUDES} are passed to @dfn{semantic-brute-find-tag-by-function}.

@obsolete{semantic-find-nonterminal-by-property,semantic-brute-find-tag-by-property} @end defun

@defun semantic-brute-find-tag-by-attribute attr streamorbuffer &optional search-parts search-includes Find all tags with a given @var{ATTR} in @var{STREAMORBUFFER}. @var{ATTR} is a symbol key into the attributes list. Optional argument @var{SEARCH-PARTS} and @var{SEARCH-INCLUDES} are passed to @dfn{semantic-brute-find-tag-by-function}.

@obsolete{semantic-find-nonterminal-by-extra-spec,semantic-brute-find-tag-by-attribute} @end defun

@defun semantic-brute-find-tag-by-attribute-value attr value streamorbuffer &optional search-parts search-includes Find all tags with a given @var{ATTR} equal to @var{VALUE} in @var{STREAMORBUFFER}. @var{ATTR} is a symbol key into the attributes list. @var{VALUE} is the value that @var{ATTR} should match. Optional argument @var{SEARCH-PARTS} and @var{SEARCH-INCLUDES} are passed to @dfn{semantic-brute-find-tag-by-function}.

@obsolete{semantic-find-nonterminal-by-extra-spec-value,semantic-brute-find-tag-by-attribute-value} @end defun

@defun semantic-brute-find-tag-by-position position streamorbuffer &optional nomedian Find a nonterminal covering @var{POSITION} within @var{STREAMORBUFFER}. @var{POSITION} is a number, or marker. If @var{NOMEDIAN} is non-@code{nil}, don't do the median calculation, and return nil.

@obsolete{semantic-find-nonterminal-by-position,semantic-brute-find-tag-by-position} @end defun

@defun semantic-brute-find-innermost-tag-by-position position streamorbuffer &optional nomedian Find a list of tags covering @var{POSITION} within @var{STREAMORBUFFER}. @var{POSITION} is a number, or marker. If @var{NOMEDIAN} is non-@code{nil}, don't do the median calculation, and return nil. This function will find the topmost item, and recurse until no more details are available of findable.

@obsolete{semantic-find-innermost-nonterminal-by-position,semantic-brute-find-innermost-tag-by-position} @end defun

@defun semantic-brute-find-tag-by-class class streamorbuffer &optional search-parts search-includes Find all tags with a class @var{CLASS} within @var{STREAMORBUFFER}. @var{CLASS} is a symbol representing the class of the tags to find. See @dfn{semantic-tag-class}. Optional argument @var{SEARCH-PARTS} and @var{SEARCH-INCLUDES} are passed to @dfn{semantic-brute-find-tag-by-function}.

Use @code{semantic-find-tag-by-class} instead.

@obsolete{semantic-find-nonterminal-by-token,semantic-brute-find-tag-by-class} @end defun

@defun semantic-brute-find-tag-standard streamorbuffer &optional search-parts search-includes Find all tags in @var{STREAMORBUFFER} which define simple class types. See @dfn{semantic-tag-class}. Optional argument @var{SEARCH-PARTS} and @var{SEARCH-INCLUDES} are passed to @dfn{semantic-brute-find-tag-by-function}.

@obsolete{semantic-find-nonterminal-standard,semantic-brute-find-tag-standard} @end defun

@defun semantic-brute-find-tag-by-type type streamorbuffer &optional search-parts search-includes Find all tags with type @var{TYPE} within @var{STREAMORBUFFER}. @var{TYPE} is a string which is the name of the type of the tags returned. See @dfn{semantic-tag-type}. Optional argument @var{SEARCH-PARTS} and @var{SEARCH-INCLUDES} are passed to @dfn{semantic-brute-find-tag-by-function}.

@obsolete{semantic-find-nonterminal-by-type,semantic-brute-find-tag-by-type} @end defun

@defun semantic-brute-find-tag-by-function function streamorbuffer &optional search-parts search-includes Find all tags for which FUNCTION's value is non-@code{nil} within @var{STREAMORBUFFER}. @var{FUNCTION} must return non-@code{nil} if an element of @var{STREAM} will be included in the new list.

If optional argument @var{SEARCH-PARTS} is non-@code{nil}, all sub-parts of tags are searched. The overloadable function @code{semantic-tag-componenets} is used for the searching child lists. If @var{SEARCH-PARTS} is the symbol @code{'positiononly}, then only children that have positional information are searched.

If @var{SEARCH-INCLUDES} is non-@code{nil}, then all include files are also searched for matches. This parameter hasn't be active for a while and is obsolete.

@obsolete{semantic-find-nonterminal-by-function,semantic-brute-find-tag-by-function} @end defun

@defun semantic-brute-find-first-tag-by-function function streamorbuffer &optional search-parts search-includes Find the first nonterminal which @var{FUNCTION} match within @var{STREAMORBUFFER}. @var{FUNCTION} must return non-@code{nil} if an element of @var{STREAM} will be included in the new list.

The following parameters were never implemented.

If optional argument @var{SEARCH-PARTS}, all sub-parts of tags are searched. The overloadable function @dfn{semantic-tag-components} is used for searching. If @var{SEARCH-INCLUDES} is non-@code{nil}, then all include files are also searched for matches.

@obsolete{semantic-find-nonterminal-by-function-first-match,semantic-brute-find-first-tag-by-function} @end defun

@node Specialty Search @section Specialty Search

There are some specialty searches needed by some semantic tools that could prove useful. These specialty searches often do not match against some single attribute as most breadth searches do.

@defun semantic-find-tags-of-compound-type &optional table @anchor{semantic-find-tags-of-compound-type} Find all tags which are a compound type in @var{TABLE}. Compound types are structures, or other data type which is not of a primitive nature, such as int or double. Used in completion. @end defun

@defun semantic-find-tags-by-scope-protection scopeprotection parent &optional table @anchor{semantic-find-tags-by-scope-protection} Find all tags accessable by @var{SCOPEPROTECTION}. @var{SCOPEPROTECTION} is a symbol which can be returned by the method @dfn{semantic-tag-protection}. @var{A} hard-coded order is used to determine a match. @var{PARENT} is a tag representing the @var{PARENT} slot needed for @dfn{semantic-tag-protection}. @var{TABLE} is a list of tags (a subset of @var{PARENT} members) to scan. If @var{TABLE} is @code{nil}, the type members of @var{PARENT} are used. See @dfn{semantic-tag-protected-p} for details on which tags are returned. @end defun

@defun semantic-find-tags-external-children-of-type type &optional table @anchor{semantic-find-tags-external-children-of-type} Find all tags in whose parent is @var{TYPE} in @var{TABLE}. These tags are defined outside the scope of the original @var{TYPE} declaration. @var{TABLE} is a tag table. See @dfn{semantic-something-to-tag-table}. @end defun

@node Custom Search @section Custom Search

The searching framework for semantic for tag tables has two basic root methods. One is a function and the other is a macro. The functional version is needed if some sort of macro conflict arrises. The macro version is useful because it eliminates a level of function call, and is faster.

@defun semantic--find-tags-by-function predicate &optional table @anchor{semantic--find-tags-by-function} Find tags for which @var{PREDICATE} is non-@code{nil} in @var{TABLE}. @var{PREDICATE} is a lambda expression which accepts on @var{TAG}. @var{TABLE} is a semantic tags table. See @dfn{semantic-something-to-tag-table}. @end defun

@defun semantic--find-tags-by-macro form &optional table @anchor{semantic--find-tags-by-macro} Find tags for which @var{FORM} is non-@code{nil} in @var{TABLE}. @var{TABLE} is a semantic tags table. See @dfn{semantic-something-to-tag-table}. @end defun

@c @defun semantic-recursive-find-nonterminal-by-name name buffer @c Recursivly find the first occurance of @var{NAME}. @c Start search with @var{BUFFER}. Recurse through all dependencies till @c found. The return item is of the form (@var{BUFFER} @var{TOKEN}) @c where @var{BUFFER} is the buffer in which @var{TOKEN} (the token found @c to match @var{NAME}) was found. @c @c @var{THIS} ISN'T @var{USED} @var{IN} @var{SEMANTIC}. @var{DELETE} @c @var{ME} @var{SOON}. @c @end defun

@node Tags at Point @chapter Tags at Point

When you need to get the tag the cursor is on, there is a more efficient mechanism than using @code{semantic-brute-find-tag-by-position}. This mechanism directly queries the overlays the parsing step leaves in the buffer. This provides for very rapid retrieval of what function or variable the cursor is currently in. @refill

These functions query the current buffer's overlay system for tags.

@defun semantic-find-tag-by-overlay &optional positionormarker buffer Find all tags covering @var{POSITIONORMARKER} by using overlays. If @var{POSITIONORMARKER} is @code{nil}, use the current point. Optional @var{BUFFER} is used if @var{POSITIONORMARKER} is a number, otherwise the current buffer is used. This finds all tags covering the specified position by checking for all overlays covering the current spot. They are then sorted from largest to smallest via the start location.

@obsolete{semantic-find-nonterminal-by-overlay,semantic-find-tag-by-overlay} @end defun

@defun semantic-find-tag-by-overlay-in-region start end &optional buffer Find all tags which exist in whole or in part between @var{START} and @var{END}. Uses overlays to determine positin. Optional @var{BUFFER} argument specifies the buffer to use.

@obsolete{semantic-find-nonterminal-by-overlay-in-region,semantic-find-tag-by-overlay-in-region} @end defun

@defun semantic-find-tag-by-overlay-next &optional start buffer Find the next tag after @var{START} in @var{BUFFER}. If @var{START} is in an overlay, find the tag which starts next, not the current tag. @end defun

@defun semantic-find-tag-by-overlay-prev &optional start buffer Find the next tag before @var{START} in @var{BUFFER}. If @var{START} is in an overlay, find the tag which starts next, not the current tag. @end defun

@defun semantic-current-tag Return the current tag in the current buffer. If there are more than one in the same location, return the smallest tag. Return @code{nil} if there is no tag here.

@obsolete{semantic-current-nonterminal,semantic-current-tag} @end defun

@defun semantic-current-tag-parent Return the current tags parent in the current buffer. @var{A} tag's parent would be a containing structure, such as a type containing a field. Return @code{nil} if there is no parent.

@obsolete{semantic-current-nonterminal-parent,semantic-current-tag-parent} @end defun

@node Tag Decoration @chapter Tag Decoration

Tags can be decorated in different ways. One way a user can control it is through @code{semantic-decoration-mode}. @inforef{semantic-user,,Tag Decoration Mode}

Applications can use the same routines to decorate tags as well.

@menu * Tag Highlighting:: Highlighting a tag * Tag Visible Properties:: Invisible, intangible and read only * Tag Secondary Overlays:: Decorating parts of a tag text * Tag Folding:: Visibly Folding up tags @end menu

@node Tag Highlighting @section Highlighting

@defun semantic-highlight-tag tag &optional face @anchor{semantic-highlight-tag} Specify that @var{tag} should be highlighted. Optional @var{face} specifies the face to use. @obsolete{semantic-highlight-token,semantic-highlight-tag} @end defun

@defun semantic-unhighlight-tag tag @anchor{semantic-unhighlight-tag} Unhighlight @var{tag}, restoring it's previous face. @obsolete{semantic-unhighlight-token,semantic-unhighlight-tag} @end defun

@defun semantic-momentary-highlight-tag tag &optional face @anchor{semantic-momentary-highlight-tag} Highlight @var{tag}, removing highlighting when the user hits a key. Optional argument @var{face} is the face to use for highlighting. If @var{face} is not specified, then @code{highlight} will be used. @obsolete{semantic-momentary-highlight-token,semantic-momentary-highlight-tag} @end defun

The highlighting routines do their work by changing the face property of the tag overlay. The raw routine is:

@defun semantic-set-tag-face tag face @anchor{semantic-set-tag-face} Specify that @var{tag} should use @var{face} for display. @obsolete{semantic-set-token-face,semantic-set-tag-face} @end defun

@node Tag Visible Properties @section Changing a tag's visible properties

You can give a tag other properties as well, such as making it invisible or intangible. You can control how code is edited programatically through these routines.

@defun semantic-set-tag-invisible tag &optional visible @anchor{semantic-set-tag-invisible} Enable the text in @var{tag} to be made invisible. If @var{visible} is non-@code{nil}, make the text visible. @obsolete{semantic-set-token-invisible,semantic-set-tag-invisible} @end defun

@defun semantic-tag-invisible-p tag @anchor{semantic-tag-invisible-p} Return non-@code{nil} if @var{tag} is invisible. @obsolete{semantic-token-invisible-p,semantic-tag-invisible-p} @end defun

@defun semantic-set-tag-intangible tag &optional tangible @anchor{semantic-set-tag-intangible} Enable the text in @var{tag} to be made intangible. If @var{tangible} is non-@code{nil}, make the text visible. This function does not have meaning in XEmacs because it seems that the extent 'intangible' property does not exist. @obsolete{semantic-set-token-intangible,semantic-set-tag-intangible} @end defun

@defun semantic-tag-intangible-p tag @anchor{semantic-tag-intangible-p} Return non-@code{nil} if @var{tag} is intangible. This function does not have meaning in XEmacs because it seems that the extent 'intangible' property does not exist. @obsolete{semantic-token-intangible-p,semantic-tag-intangible-p} @end defun

@defun semantic-set-tag-read-only tag &optional writable @anchor{semantic-set-tag-read-only} Enable the text in @var{tag} to be made read-only. Optional argument @var{writable} should be non-@code{nil} to make the text writable instead of read-only. @obsolete{semantic-set-token-read-only,semantic-set-tag-read-only} @end defun

@defun semantic-tag-read-only-p tag @anchor{semantic-tag-read-only-p} Return non-@code{nil} if the current @var{tag} is marked read only. @obsolete{semantic-token-read-only-p,semantic-tag-read-only-p} @end defun

@node Tag Secondary Overlays @section Secondary Overlays

Each tag which is being visited in a buffer has one overlay. This overlay is used to track where the tag is while a user is editing, and can also be used for fast tag identification, and some simple decoration techniques.

A secondary overlay associates a new overlay object with a tag which does not cover the entire body of a tag. You can then put visible features on that overlay to decorate portions of a tag. This is how tag boundaries are drawn.

@subsection Creation and Deletion

@defun semantic-tag-create-secondary-overlay tag &optional link-hook @anchor{semantic-tag-create-secondary-overlay} Create a secondary overlay for @var{tag}. Returns an overlay. The overlay is also saved in @var{tag}. @var{link-hook} is a function called whenever @var{tag} is to be linked into a buffer. It should take @var{tag} and @var{overlay} as arguments. The @var{link-hook} should be used to position and set properties on the generated secondary overlay. @end defun

@defun semantic-tag-delete-secondary-overlay tag overlay-or-property @anchor{semantic-tag-delete-secondary-overlay} Delete from @var{tag} the secondary overlay @var{overlay-or-property}. If @var{overlay-or-property} is an overlay, delete that overlay. If @var{overlay-or-property} is a symbol, find the overlay with that property. @end defun

@subsection Queries

@defun semantic-tag-get-secondary-overlay tag property @anchor{semantic-tag-get-secondary-overlay} Return secondary overlays from @var{tag} with @var{property}. @var{property} is a symbol and all overlays with that symbol are returned.. @end defun

@defun semantic-tag-secondary-overlays tag @anchor{semantic-tag-secondary-overlays} Return a list of secondary overlays active on @var{tag}. @end defun

@subsection Linking and Unlinking

When a tag's file is put in a buffer, that tag is ``linked'' into the buffer. When the buffer is deleted, the tag is ``unlinked''. This process adds and removes the default overlay on the tag. Secondary overlays use @code{semantic-tag-add-hook} and @code{semantic-tag-remove-hook} too apply link and unlink hooks.

This allows the secondary overlays to be automatically removed or added by the seconary overlay system whenever a tag's file goes in or out of a buffer.

@node Tag Folding @section Folding

Using secondary overlays, a set of tag folding routines are made available. These routines are similar to the tag visible properties.

@defun semantic-set-tag-folded tag &optional folded @anchor{semantic-set-tag-folded} Fold @var{tag}, such that only the first line of text is shown. Optional argument @var{folded} should be non-@code{nil} to fold the tag. @code{nil} implies the tag should be fully shown. @end defun

@defun semantic-tag-folded-p tag @anchor{semantic-tag-folded-p} Non-@code{nil} if @var{tag} is currently folded. @end defun

@node Tag Sorting @chapter Tag Sorting

Sometimes it is important to reorganize a tag stream into a form that is better for display to a user. It is important to not use functions with side effects that could effect the tag cache.

There are some existing utility functions which will reorganize the tag list for you.

@defun semantic-unique-tag-table tags @anchor{semantic-unique-tag-table} Scan a list of @var{TAGS}, removing duplicates. This must first sort the tags by position ascending. @var{TAGS} are removed only if they are equivalent, as can happen when multiple tag sources are scanned. @end defun

@defun semantic-unique-tag-table-by-name tags @anchor{semantic-unique-tag-table-by-name} Scan a list of @var{TAGS}, removing duplicate names. This must first sort the tags by name alphabetically ascending. @end defun

@defun semantic-bucketize tags &optional parent filter @anchor{semantic-bucketize} Sort @var{tags} into a group of buckets based on tag class. Unknown classes are placed in a Misc bucket. Type bucket names are defined by either @code{semantic-symbol->name-assoc-list}. If @var{parent} is specified, then @var{tags} belong to this @var{parent} in some way. This will use @code{semantic-symbol->name-assoc-list-for-type-parts} to generate bucket names. Optional argument @var{filter} is a filter function to be applied to each bucket. The filter function will take one argument, which is a list of tokens, and may re-organize the list with side-effects. @end defun

@defvar semantic-bucketize-tag-class @anchor{semantic-bucketize-tag-class} Function used to get a symbol describing the class of a tag. This function must take one argument of a semantic tag. It should return a symbol found in @code{semantic-symbol->name-assoc-list} which @dfn{semantic-bucketize} uses to bin up tokens. To create new bins for an application augment @code{semantic-symbol->name-assoc-list}, and @code{semantic-symbol->name-assoc-list-for-type-parts} in addition to setting this variable (locally in your function). @end defvar

@defun semantic-adopt-external-members tags @anchor{semantic-adopt-external-members} Rebuild @var{tags} so that externally defined members are regrouped. Some languages such as @var{c}++ and @var{clos} permit the declaration of member functions outside the definition of the class. It is easier to study the structure of a program when such methods are grouped together more logically.

This function uses @dfn{semantic-tag-external-member-p} to determine when a potential child is an externally defined member.

Note: Applications which use this function must account for token types which do not have a position, but have children which *do* have positions.

Applications should use @code{semantic-mark-external-member-function} to modify all tags which are found as externally defined to some type. For example, changing the token type for generating extra buckets with the bucket function. @end defun

@defvar semantic-orphaned-member-metaparent-type @anchor{semantic-orphaned-member-metaparent-type} In @dfn{semantic-adopt-external-members}, the type of @code{'type} for metaparents. A metaparent is a made-up type semantic token used to hold the child list of orphaned members of a named type. @end defvar

@defvar semantic-mark-external-member-function @anchor{semantic-mark-external-member-function} Function called when an externally defined orphan is found. By default, the token is always marked with the @code{adopted} property. This function should be locally bound by a program that needs to add additional behaviors into the token list. This function is called with two arguments. The first is @var{token} which is a shallow copy of the token to be modified. The second is the @var{parent} which is adopting @var{token}. This function should return @var{token} (or a copy of it) which is then integrated into the revised token list. @end defvar

@node Tag Completion @chapter Tag Completion

Often time, it is useful to ask the user for the name of some tag. This can be as simple as just prompting for a name, but often time, the semantics can be quite complex. If two tags have the same name, which do you choose?

Semantic provides a completion engine for prompting for tags by name, and providing fancy ways to display completion lists that allow the user to choose a tag if several have the same name.

To use a completion function in your interactive function, you can augment the @dfn{interactive} command like this:

@example (defun my-function (tag)
"Do something to TAG."
(interactive (list (my-completion-function "Tag: ")))
...) @end example

@menu * Tag Completion Convenience Functions:: Provided functions * Custom Tag Completion Functions:: Build your own completion function * Old Tag Completion:: Completion functions from older releases @end menu

@node Tag Completion Convenience Functions @section Tag Completion Convenience Functions

There are some pre written completion functions that can be used in your programs.

@defun semantic-complete-read-tag-buffer-deep prompt &optional default-tag initial-input history Ask for a tag by name from the current buffer. Available tags are from the current buffer, at any level. Completion options are presented in a traditional way, with highlighting to resolve same-name collisions. @var{PROMPT} is a string to prompt with. @var{DEFAULT-TAG} is a semantic tag or string to use as the default value. If @var{INITIAL-INPUT} is non-@code{nil}, insert it in the minibuffer initially. @var{HISTORY} is a symbol representing a variable to story the history in. @end defun

@defun semantic-complete-read-tag-project prompt &optional default-tag initial-input history @anchor{semantic-complete-read-tag-project} Ask for a tag by name from the current project. Available tags are from the current project, at the top level. Completion options are presented in a traditional way, with highlighting to resolve same-name collisions. @var{prompt} is a string to prompt with. @var{default-tag} is a semantic tag or string to use as the default value. If @var{initial-input} is non-@code{nil}, insert it in the minibuffer initially. @var{history} is a symbol representing a variable to store the history in. @end defun

@defun semantic-complete-read-tag-analyzer prompt &optional context history @anchor{semantic-complete-read-tag-analyzer} Ask for a tag by name based on the current context. The function @dfn{semantic-analyze-current-context} is used to calculate the context. @dfn{semantic-analyze-possible-completions} is used to generate the list of possible completions. @var{prompt} is the first part of the prompt. Additional prompt is added based on the contexts full prefix. @var{context} is the semantic analyzer context to start with. @var{history} is a symbol representing a variable to stor the history in. usually a default-tag and initial-input are available for completion prompts. these are calculated from the @var{context} variable passed in. @end defun

@defun semantic-complete-inline-analyzer context @anchor{semantic-complete-inline-analyzer} Complete a symbol name by name based on the current context. This is similar to @code{semantic-complete-read-tag-analyze}, except that the completion interaction is in the buffer where the context was calculated from. @var{context} is the semantic analyzer context to start with. See @code{semantic-complete-inline-tag-engine} for details on how completion works. @end defun

@node Custom Tag Completion Functions @section Custom Tag Completion Functions

There aren't many built in completion functions, but there are many parts that can be put together into custom completion functions.

A completion function is built up of three important parts.

@table @asis @item Tag Collection Something that selects tags, and provides some list of tags available, such as all functions, or all classes named ``bob''. @item Typing and selecting The prompt where you can type in the name of a tag. @item Displaying possible completion values A mechanism for displaying completion lists. @end table

There is one typing and selecting routine that can be used to create your custom completion prompt.

@defun semantic-complete-read-tag-engine collector displayor prompt default-tag initial-input history @anchor{semantic-complete-read-tag-engine} Read a semantic tag, and return a tag for the selection. Argument @var{collector} is an object which can be used to to calculate a list of possible hits. See @code{semantic-completion-collector-engine} for details on @var{collector}. Argumeng @var{displayor} is an object used to display a list of possible completions for a given prefix. See@code{semantic-completion-display-engine} for details on @var{displayor}. @var{prompt} is a string to prompt with. @var{default-tag} is a semantic tag or string to use as the default value. If @var{initial-input} is non-@code{nil}, insert it in the minibuffer initially. @var{history} is a symbol representing a variable to story the history in. @end defun

As you can see, this takes one @dfn{collector}, and one @dfn{displayor}. These are objects created for this prompt at runtime. The completion engine then uses to perform their tasks.

For example:

@example (defun semantic-complete-read-tag-buffer-deep (prompt &optional default-tag initial-input history)
"Ask for a tag by name from the current buffer. PROMPT is a string to prompt with. DEFAULT-TAG is a semantic tag or string to use as the default value. If INITIAL-INPUT is non-nil, insert it in the minibuffer initially. HISTORY is a symbol representing a variable to story the history in."
(semantic-complete-read-tag-engine
(semantic-collector-buffer-deep prompt :buffer (current-buffer))
(semantic-displayor-traditional-with-focus-highlight "simple")
;;(semantic-displayor-tooltip "simple")
prompt
default-tag
initial-input
history)
) @end example

@menu * Tag Collectors:: * Tag Displayors:: @end menu

@node Tag Collectors @subsection Tag Collectors

A tag collector is an object that inherits from @dfn{semantic-collector-abstract}. A new collector is needed for each specialized type of prompt that draws from a different selection of tags.

You can have a collector that satisfies multiple purposes using slots filled in the initializer for that object.

Collectors inherit from some of the following:

@deffn Type semantic-collector-abstract @anchor{semantic-collector-abstract} Root class for completion engines. The baseclass provides basic functionality for interacting with a completion displayor object, and tracking the current progress of a completion. @end deffn

@deffn Type semantic-collector-buffer-abstract @anchor{semantic-collector-buffer-abstract} Root class for per-buffer completion engines. These collectors track themselves on a per-buffer basis. @end deffn

@deffn Type semantic-collector-project-abstract @anchor{semantic-collector-project-abstract} Root class for project wide completion engines. Uses semanticdb for searching all tags in the current project. @end deffn

Available instantiable classes are:

@deffn Type semantic-collector-buffer-deep @anchor{semantic-collector-buffer-deep} Completion engine for tags in the current buffer. When searching for a tag, uses semantic deep searche functions. Basics search only in the current buffer. @end deffn

@deffn Type semantic-collector-project @anchor{semantic-collector-project} Completion engine for tags in a project. @end deffn

@deffn Type semantic-collector-project-brutish @anchor{semantic-collector-project-brutish} Completion engine for tags in a project. @end deffn

@deffn Type semantic-collector-analyze-completions @anchor{semantic-collector-analyze-completions} Completion engine that uses the context analyzer to provide options. The only options available for completion are those which can be logically inserted into the current context. @end deffn

NOTE: Add sections for writing new collectors.

@c (eieiodoc-class semantic-collector-abstract "cc")

@c END OF COLLECTOR

@node Tag Displayors @subsection Tag Displayors

When a user is interacting with the prompt, and requests completion, those tags are drawn from the collector. If the user asks for a list of completion by hitting a complete key twice, then the list of completions heeds to be displayed in some way.

Displayors can be used to manage the display of all tags currently available, AND often needs to be used to @b{focus} one one particular tag of many in a visible way.

All displayors inherit from the displayor baseclass that defines the default behaviors:

@deffn Type semantic-displayor-abstract @anchor{semantic-displayor-abstract} Manages the display of some number of tags. Provides the basics for a displayor, including interacting with a collector, and tracking tables of completion to display. @end deffn

@deffn Type semantic-displayor-focus-abstract @anchor{semantic-displayor-focus-abstract} A displayor which has the ability to focus in on one tag. Focusing is a way of differentiationg between multiple tags which have the same name. @end deffn

Distinct implementations of displayors include:

@deffn Type semantic-displayor-traditional @anchor{semantic-displayor-traditional} Traditional display mechanism for a list of possible completions. Completions are showin in a new buffer and listed with the ability to click on the items to aid in completion. @end deffn

@deffn Type semantic-displayor-traditional-with-focus-highlight @anchor{semantic-displayor-traditional-with-focus-highlight} A traditional displayor which can focus on a tag by showing it. Same as @code{semantic-displayor-traditional}, but with selection between multiple tags with the same name done by 'focusing' on the source location of the different tags to differentiate them.

@end deffn

@deffn Type semantic-displayor-tooltip @anchor{semantic-displayor-tooltip} Display mechanism using tooltip for a list of possible completions. @end deffn

NOTE: Add sections for writing new collectors.

@c (eieiodoc-class semantic-displayor-abstract "cd")

@c END OF DISPLAYOR

@node Old Tag Completion @section Older Tag Completion functions

These are older completion functions. They may still be useful.

@defun semantic-read-symbol prompt &optional default stream filter @anchor{semantic-read-symbol} Read a symbol name from the user for the current buffer. @var{prompt} is the prompt to use. Optional arguments: @var{default} is the default choice. If no default is given, one is read from under point. @var{stream} is the list of tokens to complete from. @var{filter} is provides a filter on the types of things to complete. @var{filter} must be a function to call on each element. @end defun

@defun semantic-read-variable prompt &optional default stream @anchor{semantic-read-variable} Read a variable name from the user for the current buffer. @var{prompt} is the prompt to use. Optional arguments: @var{default} is the default choice. If no default is given, one is read from under point. @var{stream} is the list of tokens to complete from. @end defun

@defun semantic-read-function prompt &optional default stream @anchor{semantic-read-function} Read a function name from the user for the current buffer. @var{prompt} is the prompt to use. Optional arguments: @var{default} is the default choice. If no default is given, one is read from under point. @var{stream} is the list of tags to complete from. @end defun

@defun semantic-read-type prompt &optional default stream @anchor{semantic-read-type} Read a type name from the user for the current buffer. @var{prompt} is the prompt to use. Optional arguments: @var{default} is the default choice. If no default is given, one is read from under point. @var{stream} is the list of tags to complete from. @end defun

@node Override Methods @chapter Override Methods

These functions are called `override methods' because they provide generic behaviors, which a given language can override. For example, finding a dependency file in Emacs lisp can be done with the `locate-library' command (which overrides the default behavior.) In C, a dependency can be found by searching a generic search path which can be passed in via a variable.

If you plan to use one of these functions from a buffer that is not of the same major-mode as the original tag, you can use this form to make sure the correct action takes place:

@defun semantic-with-mode-bindings mode &rest body @anchor{semantic-with-mode-bindings} Evaluate @var{BODY} with the local bindings of @var{MODE}. The current mode bindings are saved, @var{BODY} is evaluated, and the saved bindings are restored, even in case of an abnormal exit. Value is what @var{BODY} returns. @end defun

For more on override methods, @inforef{Semantic Overload Mechanism, , lang-support-guide}.

@menu * Format Tag:: Converting Tokens into text strings * Tag Members:: Tags in tags * Tag Details:: Arbitrary token detail fetching * Making New Methods:: How to add your own methods for a tool @end menu

@node Format Tag @section Format Tag

Any given tag consists of Meta information which is best viewed in some textual form. This could be as simple as the tag's name, or as a prototype to be added to header file in C. Not only are there several default converters from a Tag into text, but there is also some convenient variables that can be used with them. Use these variables to allow options on output forms when displaying tags in your programs.

@defvar semantic-format-tag-functions @anchor{semantic-format-tag-functions} List of functions which convert a tag to text. Each function must take the parameters @var{tag} &optional @var{parent} @var{color}. @var{tag} is the tag to convert. @var{parent} is a parent tag or name which refers to the structure or class which contains @var{tag}. @var{parent} is @var{not} a class which a @var{tag} would claim as a parent. @var{color} indicates that the generated text should be colored using @code{font-lock}. @end defvar

@defvar semantic-format-tag-custom-list @anchor{semantic-format-tag-custom-list} A List used by customizeable variables to choose a tag to text function. Use this variable in the @code{:type} field of a customizable variable. @end defvar

Every tag to text conversion function must take the same parameters, which are @var{TAG}, the tag to be converted, @var{PARENT}, the containing parent (like a structure which contains a variable), and @var{COLOR}, which is a flag specifying that color should be applied to the returned string.

When creating, or using these strings, particularly with color, use @dfn{concat} to build up larger strings instead of @dfn{format}. This will preserve text properties.

@defun semantic-format-tag-name tag &optional parent color @anchor{semantic-format-tag-name} Return the name string describing @var{tag}. The name is the shortest possible representation. Optional argument @var{parent} is the parent type if @var{tag} is a detail. Optional argument @var{color} means highlight the prototype with font-lock colors. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-name-nonterminal,semantic-format-tag-name} @end defun

@defun semantic-format-tag-abbreviate tag &optional parent color @anchor{semantic-format-tag-abbreviate} Return an abbreviated string describing @var{tag}. The abbreviation is to be short, with possible symbols indicating the type of tag, or other information. Optional argument @var{parent} is the parent type if @var{tag} is a detail. Optional argument @var{color} means highlight the prototype with font-lock colors. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-abbreviate-nonterminal,semantic-format-tag-abbreviate} @end defun

@defun semantic-format-tag-summarize tag &optional parent color @anchor{semantic-format-tag-summarize} Summarize @var{tag} in a reasonable way. Optional argument @var{parent} is the parent type if @var{tag} is a detail. Optional argument @var{color} means highlight the prototype with font-lock colors. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-summerize-nonterminal,semantic-format-tag-summarize} @end defun

@defun semantic-format-tag-prototype tag &optional parent color @anchor{semantic-format-tag-prototype} Return a prototype for @var{tag}. This function should be overloaded, though it need not be used. This is because it can be used to create code by language independent tools. Optional argument @var{parent} is the parent type if @var{tag} is a detail. Optional argument @var{color} means highlight the prototype with font-lock colors. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-prototype-nonterminal,semantic-format-tag-prototype} @end defun

@defun semantic-format-tag-concise-prototype tag &optional parent color @anchor{semantic-format-tag-concise-prototype} Return a concise prototype for @var{tag}. Optional argument @var{parent} is the parent type if @var{tag} is a detail. Optional argument @var{color} means highlight the prototype with font-lock colors. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-concise-prototype-nonterminal,semantic-format-tag-concise-prototype} @end defun

@defun semantic-format-tag-uml-abbreviate tag &optional parent color @anchor{semantic-format-tag-uml-abbreviate} Return a @var{uml} style abbreviation for @var{tag}. Optional argument @var{parent} is the parent type if @var{tag} is a detail. Optional argument @var{color} means highlight the prototype with font-lock colors. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-uml-abbreviate-nonterminal,semantic-format-tag-uml-abbreviate} @end defun

@defun semantic-format-tag-uml-prototype tag &optional parent color @anchor{semantic-format-tag-uml-prototype} Return a @var{uml} style prototype for @var{tag}. Optional argument @var{parent} is the parent type if @var{tag} is a detail. Optional argument @var{color} means highlight the prototype with font-lock colors. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-uml-prototype-nonterminal,semantic-format-tag-uml-prototype} @end defun

@defun semantic-format-tag-uml-concise-prototype tag &optional parent color @anchor{semantic-format-tag-uml-concise-prototype} Return a @var{uml} style concise prototype for @var{tag}. Optional argument @var{parent} is the parent type if @var{tag} is a detail. Optional argument @var{color} means highlight the prototype with font-lock colors. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-uml-concise-prototype-nonterminal,semantic-format-tag-uml-concise-prototype} @end defun

@defun semantic-format-tag-prin1 tag &optional parent color @anchor{semantic-format-tag-prin1} Convert @var{tag} to a string that is the print name for @var{tag}. @var{parent} and @var{color} are ignored. @obsolete{semantic-prin1-nonterminal,semantic-format-tag-prin1} @end defun

An additional utility will return a string for just the data type of a tag. This function is used in the above routines as well.

@defun semantic-format-tag-type tag color @anchor{semantic-format-tag-type} Convert the data type of @var{tag} to a string usable in tag formatting. It is presumed that @var{type} is a string or semantic tag. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@node Tag Members @section Tag Members

Tags are often in a hierarchical form, meaning that a tag found in the top-level list may itself contain additional members.

The following overridable functions can fetch those tags.

@defun semantic-tag-components tag Return a list of components for @var{tag}. A Component is a part of @var{tag} which itself may be a @var{tag}. Examples include the elements of a structure in a tag of class `type, or the list of arguments to a tag of class @code{'function}. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-tag-components-with-overlays tag Return the list of top level components belonging to @var{tag}. Children are any sub-tags which contain overlays.

Default behavior is to get @dfn{semantic-tag-components} in addition to the components of an anonymous types (if applicable.)

Note for language authors:
If a mode defines a language tag that has tags in it with overlays you should still return them with this function. Ignoring this step will prevent several features from working correctly. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

Some languages can define parts of a tag outside the actual scope of the parent tag. You can fetch information about them with these overload functions.

@defun semantic-tag-external-member-p parent tag @anchor{semantic-tag-external-member-p} Return non-@code{nil} if @var{parent} is the parent of @var{tag}. @var{tag} is an external member of @var{parent} when it is somehow tagged as having @var{parent} as it's parent. @var{parent} and @var{tag} must both be semantic tags.

The default behavior, if not overridden with @code{tag-external-member-p} is to match @code{:parent} attribute in the name of @var{tag}.

If this function is overridden, use @code{semantic-tag-external-member-children-p-default} to also include the default behavior, and merely extend your own. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-nonterminal-external-member-p,semantic-tag-external-member-p} @end defun

@defun semantic-tag-external-member-children tag &optional usedb @anchor{semantic-tag-external-member-children} Return the list of children which are not *in* @var{tag}. If optional argument @var{usedb} is non-@code{nil}, then also search files in the Semantic Database. If @var{usedb} is a list of databases, search those databases.

Children in this case are functions or types which are members of @var{tag}, such as the parts of a type, but which are not defined inside the class. @var{c}++ and @var{clos} both permit methods of a class to be defined outside the bounds of the class' definition.

The default behavior, if not overridden with @code{tag-external-member-children} is to search using @dfn{semantic-tag-external-member-p} in all top level definitions with a parent of @var{tag}.

If this function is overridden, use @dfn{semantic-tag-external-member-children-default} to also include the default behavior, and merely extend your own. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-nonterminal-external-member-children,semantic-tag-external-member-children} @end defun

@node Tag Details @section Tag Details

These functions help derive information about tags that may not be obvious for non-traditional languages with their own token types.

@defun semantic-tag-protection tag &optional parent @anchor{semantic-tag-protection} Return protection information about @var{tag} with optional @var{parent}. This function returns on of the following symbols:
@code{nil} - No special protection. Language dependent.
@code{'public} - Anyone can access this @var{tag}.
@code{'private} - Only methods in the local scope can access @var{tag}.
@code{'protected} - Like private for outside scopes, like public for child
classes. Some languages may choose to provide additional return symbols specific to themselves. Use of this function should allow for this.

The default behavior (if not overridden with @code{tag-protection} is to return a symbol based on type modifiers. This function can be overloaded (see @dfn{define-mode-local-override} for details). It makes the overload @dfn{semantic-nonterminal-protection} obsolete. @obsolete{semantic-nonterminal-protection,semantic-tag-protection} @end defun

@defun semantic-tag-protected-p tag protection &optional parent @anchor{semantic-tag-protected-p} Non-@code{nil} if @var{tag} is is protected. @var{protection} is a symbol which can be returned by the method @dfn{semantic-tag-protection}. @var{parent} is the parent data type which contains @var{tag}.

For these PROTECTIONs, true is returned if @var{tag} is: @table @asis @item @code{nil}
Always true @item private
True if @code{nil}. @item protected
True if private or @code{nil}. @item public
True if private, protected, or @code{nil}. @end table @end defun

@defun semantic-tag-abstract-p tag &optional parent @anchor{semantic-tag-abstract-p} Return non @code{nil} if @var{tag} is abstract. Optional @var{parent} is the parent tag of @var{tag}. In @var{uml}, abstract methods and classes have special meaning and behavior in how methods are overridden. In @var{uml}, abstract methods are italicized.

The default behavior (if not overridden with @code{tag-abstract-p} is to return true if @code{abstract} is in the type modifiers. This function can be overloaded (see @dfn{define-mode-local-override} for details). It makes the overload @dfn{semantic-nonterminal-abstract} obsolete. @obsolete{semantic-nonterminal-abstract,semantic-tag-abstract-p} @end defun

@defun semantic-tag-leaf-p tag &optional parent @anchor{semantic-tag-leaf-p} Return non @code{nil} if @var{tag} is leaf. Optional @var{parent} is the parent tag of @var{tag}. In @var{uml}, leaf methods and classes have special meaning and behavior.

The default behavior (if not overridden with @code{tag-leaf-p} is to return true if @code{leaf} is in the type modifiers. This function can be overloaded (see @dfn{define-mode-local-override} for details). It makes the overload @dfn{semantic-nonterminal-leaf} obsolete. @obsolete{semantic-tag-leaf,semantic-tag-leaf-p} @end defun

@defun semantic-tag-static-p tag &optional parent @anchor{semantic-tag-static-p} Return non @code{nil} if @var{tag} is static. Optional @var{parent} is the parent tag of @var{tag}. In @var{uml}, static methods and attributes mean that they are allocated in the parent class, and are not instance specific. @var{uml} notation specifies that @var{static} entries are underlined. This function can be overloaded (see @dfn{define-mode-local-override} for details). @obsolete{semantic-tag-static,semantic-tag-static-p} @end defun

@defvar semantic-dependency-include-path @anchor{semantic-dependency-include-path} Defines the include path used when searching for files. This should be a list of directories to search which is specific to the file being included.

If @dfn{semantic-dependency-tag-file} is overridden for a given language, this path is most likely ignored.

This function, reguardless of being overriden, caches the located dependency file location in the tag property @code{dependency-file}. If you override this function, you do not need to implement your own cache. Each time the buffer is fully reparsed, the cache will be reset.

@var{todo}: use ffap.el to locate such items. @end defvar

@defun semantic-dependency-tag-file &optional tag @anchor{semantic-dependency-tag-file} Find the filename represented from @var{tag}. Depends on @code{semantic-dependency-include-path} for searching. Always searches @code{.} first, then searches additional paths. This function can be overloaded (see @dfn{define-mode-local-override} for details). It makes the overload @dfn{semantic-find-dependency} obsolete. @obsolete{semantic-find-dependency,semantic-dependency-tag-file} @end defun

@defun semantic-prototype-file buffer @anchor{semantic-prototype-file} Return a file in which prototypes belonging to @var{buffer} should be placed. Default behavior (if not overridden) looks for a token specifying the prototype file, or the existence of an @var{ede} variable indicating which file prototypes belong in. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-go-to-tag tag &optional parent @anchor{semantic-go-to-tag} Go to the location of @var{tag}. @var{tag} may be a stripped element, in which case @var{parent} specifies a parent tag that has position information. Different behaviors are provided depending on the type of tag. For example, dependencies (includes) will seek out the file that is depended on (see @dfn{semantic-dependency-tag-file}. This function can be overloaded (see @dfn{define-mode-local-override} for details). It makes the overload @dfn{semantic-find-nonterminal} obsolete. @obsolete{semantic-find-nonterminal,semantic-go-to-tag} @end defun

@defun semantic-texi-find-documentation name &optional type @anchor{semantic-texi-find-documentation} Find the function or variable @var{name} of @var{type} in the texinfo source. @var{name} is a string representing some functional symbol. @var{type} is a string, such as ``variable`` or ''Command'' used to find the correct definition in case @var{name} qualifies as several things. When this function exists, @var{point} is at the definition. If the doc was not found, an error is thrown. Note: @var{type} not yet implemented. @end defun

@node Making New Methods @section Making New Methods

To create your own application specific mode local function, you should use @dfn{define-overload}. This function creates the framework needed for different mode-authors to customize your applications.

Once a function has been defined as mode-local, you should use @dfn{define-mode-local-override} to add a mode specific feature.

@defun define-overload name args docstring &rest body @anchor{define-overload} Define a new function, as with @dfn{defun} which can be overloaded. @var{name} is the name of the function to create. @var{args} are the arguments to the function. @var{docstring} is a documentation string to describe the function. The docstring will automatically had details about its overload symbol appended to the end. @var{body} is code that would be run when there is no override defined. The default is to call the function @code{NAME-default} with the appropriate arguments.

@var{body} can also include an override form that specifies which part of @var{body} is specifically overridden. This permits to specify common code run for both default and overridden implementations. An override form is one of:

@var{1}. (:override [@var{overbody}])
@var{2}. (:override-with-args @var{overargs} [@var{overbody}])

@var{overbody} is the code that would be run when there is no override defined. The default is to call the function @code{NAME-default} with the appropriate arguments deduced from @var{args}. @var{overargs} is a list of arguments passed to the override and @code{NAME-default} function, in place of those deduced from @var{args}. @end defun

@defun define-mode-local-override name mode args docstring &rest body @anchor{define-mode-local-override} Define a mode specific override of the function overload @var{name}. Has meaning only if @var{name} has been created with @dfn{define-overload}. @var{mode} is the major mode this override is being defined for. @var{args} are the function arguments, which should match those of the same named function created with @dfn{define-overload}. @var{docstring} is the documentation string. @var{body} is the implementation of this function. @obsolete{define-mode-overload-implementation,define-mode-local-override} @end defun

@node Parser Features @chapter Parser Features

If you write a program that uses a tag table in a persistent display or database, it is necessary to know when tag tables change so that your displays can be updated. This is especially important as tags can be replaced, changed, or deleted, and the associated overlays will then throw errors when you try to use them. Complete integration with tag changes can be achieved via several hooks.

If you write an application that frequenly accesses the tags tables, it is important to make sure those tags are up to date, and to make sure you application does not adversely effect all the other minor modes that may be running.

The semantic parser has many features and hooks that applications can use to provide a good user experience.

@menu * Editing Buffers:: Let the parser know you are changing a buffer. * Parser State:: Knowing if the tag table is out of date * Parser Hooks:: Knowing when tags change * Lexical Safety:: Preventing lexical errors from making a mess @end menu

@node Editing Buffers @section Editing Buffers

One interesting way to interact with the parser is to let it know that changes you are going to make will not require re-parsing.

@defvar semantic-edits-are-safe When non-@code{nil}, modifications do not require a reparse. This prevents tokens from being marked dirty, and it prevents top level edits from causing a cache check. Use this when writing programs that could cause a full reparse, but will not change the tag structure, such as adding or updating top-level comments. @end defvar

@node Parser State @section Parser State

It is sometimes useful to know what the current parsing state is. These function can let you know what level of re-parsing may be needed. Careful choices on when to reparse can make your program much faster.

@defun semantic-parse-tree-needs-update-p @anchor{semantic-parse-tree-needs-update-p} Return non-@code{nil} if the current parse tree needs to be updated. @end defun

@defun semantic-parse-tree-needs-rebuild-p @anchor{semantic-parse-tree-needs-rebuild-p} Return non-@code{nil} if the current parse tree needs to be rebuilt. @end defun

@defun semantic-parse-tree-unparseable-p @anchor{semantic-parse-tree-unparseable-p} Return non-@code{nil} if the current buffer has been marked unparseable. @end defun

@defun semantic-parse-tree-up-to-date-p @anchor{semantic-parse-tree-up-to-date-p} Return non-@code{nil} if the current parse tree is up to date. @end defun

@node Parser Hooks @section Parser Hooks

If you just want to know when a buffer changes, use this hook.

@defvar semantic-after-toplevel-cache-change-hook @anchor{semantic-after-toplevel-cache-change-hook} Hooks run after the buffer tag list has changed. This list will change when a buffer is reparsed, or when the tag list in a buffer is cleared. It is *@var{not}* called if the current tag list is partially reparsed.

Hook functions must take one argument, which is the new list of tags associated with this buffer.

For language specific hooks, make sure you define this as a local hook. @end defvar

If you want tighter interaction with how the user is editing different tags, you can use this hook instead.

@defvar semantic-after-partial-cache-change-hook @anchor{semantic-after-partial-cache-change-hook} Hooks run after the buffer cache has been updated.

This hook will run when the cache has been partially reparsed. Partial reparses are incurred when a user edits a buffer, and only the modified sections are rescanned.

Hook functions must take one argument, which is the list of tokens updated in the current buffer.

For language specific hooks, make sure you define this as a local hook. @end defvar

It is also useful to clean up any data your application is using when semantic flushes its tags table.

@defvar semantic-before-toplevel-cache-flush-hook @anchor{semantic-before-toplevel-cache-flush-hook} Hooks run before the toplevel nonterminal cache is flushed. For language specific hooks, make sure you define this as a local hook. This hook is called before a corresponding @code{semantic-after-toplevel-cache-change-hook} which is also called during a flush when the cache is given a new value of @code{nil}. @end defvar

@node Lexical Safety @section Lexical Safety

If you application frequenly requests lists of tags upon user request, it may be important to avoid lexical problems that frequenly occur when the user has partially written an expression, such as starting a string, or argument list.

You can protect your code from lexical problems with this macro:

@defun semantic-lex-catch-errors symbol &rest forms @anchor{semantic-lex-catch-errors} Using @var{SYMBOL}, execute @var{FORMS} catching lexical errors. If @var{FORMS} results in a call to the parser that throws a lexical error, the error will be caught here without the buffer's cache being thrown out of date. If there is an error, the syntax that failed is returned. If there is no error, then the last value of @var{FORMS} is returned. @end defun

It is important to provide a good @var{SYMBOL} so that these macros can nest correctly.

If you want your code to run anyway, even if there is a lexical error, using this macro like this:

@example (semantic-lex-catch-errors
(semantic-fetch-tags)) @end example

will put the parser into the 'unparseable' state, and allow other routines to request the tag table without incurring additional parser attempts.

@node Semantic Database @chapter Semantic Database

Semanticdb is a database mechanism for storing tags parsed by @semantic{}. The database operates in the background, saving tags as they are parsed between sessions. When a file is read in, and there is a previously created set of tags available for it, sematnicdb will save time by not parsing the file, and using the cached copy.

In applications, semanticdb can provide access to the sum of all tags in a project or in the basic system. This database can they be searched using a set of special routines.

@menu * Semanticdb in Programs:: Basic usage. * Semanticdb Tag Queries:: Searching for tokens in the databases. * System Databases:: Special kinds of databases for system tags. @end menu

@node Semanticdb in Programs @section Semanticdb in Programs::

If you write a program using semanticdb, you will probably want to make sure it is active in your program.

@defun semanticdb-minor-mode-p @anchor{semanticdb-minor-mode-p} Return non-@code{nil} if @code{semanticdb-minor-mode} is active. @end defun

Since semanticdb is optional, it is best if a program can gracefully degrade service when semanticdb is not available, or to throw an error letting the user know it is required to be active.

At the simplest level, you can ask if a given file is in the database, recieving a tag table. Semanticdb will give you an accurate set of tags just by asking.

@defun semanticdb-file-stream file @anchor{semanticdb-file-stream} Return a list of tags belonging to @var{file}. If file has database tags available in the database, return them. If file does not have tags available, then load the file, and create them. @end defun

Alternately, you can get at the table object for a file by asking for it.

@defun semanticdb-file-table-object file &optional dontload @anchor{semanticdb-file-table-object} Return a semanticdb table belonging to @var{file}. If file has database tags available in the database, return it. If file does not have tags available, and @var{dontload} is @code{nil}, then load the tags for @var{file}, and create a new table object for it. @var{dontload} does not affect the creation of new database objects. @end defun

@node Semanticdb Tag Queries @section Semanticdb Tag Queries

You can search for tags in the semantic database using the semanticdb-find API. It is important to note that database search functions do not return a plain list of tags. This is because some tags may not be loaded in a buffer, which means that the found tag would not have an overlay, and no way to determine where it came from.

As such, all search functions return a special Database Results list.

There are several types of searches, divided into groups by implementation. While it is possible to add new types of searches, or write custom searches, the built in searches are usually the only ones available in system backends @ref{System Databases}.

When the term @b{brute} or @b{brutish} is used as a search criteria, that is distinguishing between an include-path based search, and a search that scans everything available in a project.

Non-brute force searches assume that all symbols available in a given buffer are on the search path, or in some file that has been included, imported, or otherwise indicated in the source file itself. While not always true for interpreted languages (such as Emacs Lisp), it is common among declaritive languages.

Sometimes a brute force approach is needed, scanning every file available to the database. You may want to do this if your application is collecting data unrelated to a file currently being worked on.

@menu * DB Results:: Accessing the results of a search. * DB Search Paths:: The list of tables to search. * DB Basic Name Search:: Searches based on name. * DB Basic Brute Search:: Searches on common tag attributes. * DB Advanced Search:: Complex searches on associations * DB Generic Brute Search:: Do It Yourself search criteria @end menu

@node DB Results @subsection DB Results

The successful results of a search returns a special list of the following form:

@example
( (DATABASE TAG1 TAG2 ...) (DATABASE2 TAG3 TAG4 ...) ...) @end example

It should not be necessary to access the results in this way, however, as there are several routines that can be used to access this list.

To turn a semanticdb search result into a simple tag table, use:

@defun semanticdb-strip-find-results results &optional find-file-match @anchor{semanticdb-strip-find-results} Strip a semanticdb search @var{results} to exclude objects. This makes it appear more like the results of a @code{semantic-find-} call. Optional @var{find-file-match} loads all files associated with @var{results} into buffers. This has the side effect of enabling @dfn{semantic-tag-buffer} to return a value. @end defun

To write a function that accepts a tag table, or a semanticdb search result, use this to test if it is a semanticdb search result:

@defun semanticdb-find-results-p resultp @anchor{semanticdb-find-results-p} Non-@code{nil} if @var{resultp} is in the form of a semanticdb search result. This query only really tests the first entry in the list that is @var{resultp}, but should be good enough for debugging assertions. @end defun

@defun semanticdb-find-result-with-nil-p resultp @anchor{semanticdb-find-result-with-nil-p} Non-@code{nil} of @var{resultp} is in the form of a semanticdb search result. @code{nil} is a valid value where a @var{table} usually is, but only if the @var{tag} results include overlays. This query only really tests the first entry in the list that is @var{resultp}, but should be good enough for debugging assertions. @end defun

To operate on the search results as though it were a simple tags table, or plain list, use these routines.

@defun semanticdb-find-result-length result @anchor{semanticdb-find-result-length} Number of tags found in @var{result}. @end defun

@defun semanticdb-find-result-nth result n @anchor{semanticdb-find-result-nth} In @var{result}, return the Nth search result. This is a @var{0} based search result, with the first match being element @var{0}.

The returned value is a cons cell: (@var{tag} . @var{table}) where @var{tag} is the tag at the Nth position. @var{table} is the semanticdb table where the @var{tag} was found. Sometimes @var{table} can be @code{nil}. @end defun

@defun semanticdb-find-result-nth-in-buffer result n @anchor{semanticdb-find-result-nth-in-buffer} In @var{result}, return the Nth search result. Like @dfn{semanticdb-find-result-nth}, except that only the @var{tag} is returned, and the buffer it is found it will be made current. If the result tag has no position information, the originating buffer is still made current. @end defun

@node DB Search Paths @subsection DB Search Paths

For searches based on an include path (non-brutish) a path of tables needs to be generated. Searching a lot of tables is slow, which is why a brutish search is not always recommended. An include-based approach can also generate a lot of tables, so you can control how detailed a search is with a throttle variable.

Ideally, each language mode will have a mode-specific value for this throttle value. A user can also specify their own values if the default is not good enough.

@defvar semanticdb-find-default-throttle @anchor{semanticdb-find-default-throttle} The default throttle for @code{semanticdb-find} routines. The throttle controls how detailed the list of database tables is for a symbol lookup. The value is a list with the following keys:

@table @code @item file The file the search is being performed from. This option is here for completeness only, and is assumed to always be on. @item local Tables from the same local directory are included. This includes files directly referenced by a file name which might be in a different directory. @item project Tables from the same local project are included If @code{project} is specified, then @code{local} is assumed. @item unloaded If a table is not in memory, load it. If it is not cached on disk either, get the source, parse it, and create the table. @item system Tables from system databases. These are specifically tables from system header files, or language equivalent. @item recursive For include based searches, includes tables referenced by included files. @item omniscience Included system databases which are omniscience, or somehow know everything. Omniscience databases are found in @code{semanticdb-project-system-databases}. The Emacs Lisp system @var{db} is an omniscience database. @end table

@end defvar

You can use the command @code{semanticdb-find-test-translate-path} to interactively test out how the path translator is working. The path translation routines are:

@defun semanticdb-find-translate-path path brutish @anchor{semanticdb-find-translate-path} Translate @var{path} into a list of semantic tables. Path translation involves identifying the @var{path} input argument in one of the following ways:
@code{nil} - Take the current buffer, and use it's include list
buffer - Use that buffer's include list.
filename - Use that file's include list. If the file is not
in a buffer, see of there is a semanticdb table for it. If
not, read that file into a buffer.
tag - Get that tag's buffer of file file. See above.
table - Search that table, and it's include list.
find result - Search the results of a previous find.

In addition, once the base path is found, there is the possibility of each added table adding yet more tables to the path, so this routine can return a lengthy list.

If argument @var{brutish} is non-@code{nil}, then instead of using the include list, use all tables found in the parent project of the table identified by translating @var{path}. Such searches use brute force to scan every available table.

The return value is a list of objects of type @dfn{semanticdb-table} or it's children. In the case of passing in a find result, the result is returned unchanged.

This routine uses @dfn{semanticdb-find-table-for-include} to translate specific include tags into a semanticdb table. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semanticdb-find-table-for-include includetag &optional table @anchor{semanticdb-find-table-for-include} For a single @var{includetag} found in @var{table}, find a @dfn{semanticdb-table} object @var{includetag} is a semantic @var{tag} of class @code{'include}. @var{table} as defined by @dfn{semantic-something-to-tag-table} to identify where the tag came from. @var{table} is optional if @var{includetag} has an overlay of @code{:filename} attribute. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@node DB Basic Name Search @subsection DB Basic Name Search

These searches scan a database table collection for tags based on name. They are divided into normal and deep searches. A deep search, as with in buffer tag scanning, implies that all entries are scanned, including those in type declarations.

Normal Searches:

@defun semanticdb-find-tags-by-name name &optional path find-file-match @anchor{semanticdb-find-tags-by-name} Search for all tags matching @var{name} on @var{path}. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. @end defun

@defun semanticdb-find-tags-by-name-regexp regexp &optional path find-file-match @anchor{semanticdb-find-tags-by-name-regexp} Search for all tags matching @var{regexp} on @var{path}. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. @end defun

@defun semanticdb-find-tags-for-completion prefix &optional path find-file-match @anchor{semanticdb-find-tags-for-completion} Search for all tags matching @var{prefix} on @var{path}. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. @end defun

@defun semanticdb-find-tags-by-class class &optional path find-file-match @anchor{semanticdb-find-tags-by-class} Search for all tags of @var{class} on @var{path}. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. @end defun

Deep Searches:

@defun semanticdb-deep-find-tags-by-name name &optional path find-file-match @anchor{semanticdb-deep-find-tags-by-name} Search for all tags matching @var{name} on @var{path}. Search also in all components of top level tags founds. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. @end defun

@defun semanticdb-deep-find-tags-by-name-regexp regexp &optional path find-file-match @anchor{semanticdb-deep-find-tags-by-name-regexp} Search for all tags matching @var{regexp} on @var{path}. Search also in all components of top level tags founds. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. @end defun

@defun semanticdb-deep-find-tags-for-completion prefix &optional path find-file-match @anchor{semanticdb-deep-find-tags-for-completion} Search for all tags matching @var{prefix} on @var{path}. Search also in all components of top level tags founds. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. @end defun

@node DB Basic Brute Search @subsection DB Basic Brute Search

These searches allow searching on specific attributes of tags,
such as name, type, or other attribute.

@defun semanticdb-brute-deep-find-tags-by-name name &optional path find-file-match @anchor{semanticdb-brute-deep-find-tags-by-name} Search for all tags matching @var{name} on @var{path}. See @dfn{semanticdb-find-translate-path} for details on @var{path}. The argument @var{brutish} will be set so that searching includes all tables in the current project. @var{find-file-match} indicates that any time a matchi is found, the file associated wit that tag should be loaded into a buffer. @end defun

@node DB Advanced Search @subsection DB Advanced Search

These are searches that were needed to accomplish some
specialized tasks as discovered in utilities. Advanced searches
include matching methods defined outside some parent class.

The reason for advanced searches are so that external
repositories such as the Emacs obarray, or java @file{.class} files can
quickly answer these needed questions without dumping the entire
symbol list into Emacs for a regular semanticdb search.

@defun semanticdb-find-tags-external-children-of-type type &optional path find-file-match @anchor{semanticdb-find-tags-external-children-of-type} Search for all tags defined outside of @var{type} w/ @var{type} as a parent. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. @end defun

@node DB Generic Brute Search @subsection DB Generic Brute Search

The generic search, @dfn{semanticdb-find-nonterminal-by-function}
will call your function with every tag available. (Very slow for
system databases.)

NOTE: There is no such function. Hopefully we can get away without implementing one, which will make system databases more useful. If you are really stuck, you can use the following, though I don't recommend it.

@defun semanticdb-find-tags-collector function &optional path find-file-match brutish @anchor{semanticdb-find-tags-collector} Search for all tags returned by @var{function} over @var{path}. See @dfn{semanticdb-find-translate-path} for details on @var{path}. @var{find-file-match} indicates that any time a match is found, the file associated with that tag should be loaded into a buffer. If optional argument @var{brutish} is non-@code{nil}, then ignore include statements, and search all tables in this project tree. @end defun

@node System Databases @section System Databases

System Databases are special implementations of the database and table API which make some external tag source appear as though it were a normal buffer with semantic parsed tags available.

Search routines for these databases return a special type of table not associated with a file. It is important to be aware of this possible return value.

@deffn Type semanticdb-project-database-emacs-lisp @anchor{semanticdb-project-database-emacs-lisp} Database representing Emacs core. @end deffn

This Emacs database is loaded automatically.

@deffn Type semanticdb-project-database-ebrowse @anchor{semanticdb-project-database-ebrowse} Semantic Database deriving tags using the @var{ebrowse} tool. @var{ebrowse} is a C/C++ parser for use with @code{ebrowse} Emacs program. @end deffn

To create new EBROWSE project databases, use:

@deffn Command semanticdb-create-ebrowse-database dir @anchor{semanticdb-create-ebrowse-database} Create an @var{ebrose} database for directory @var{dir}. The database file is stored in ~/.semanticdb, or whichever directory is specified by @code{semanticdb-default-system-save-directory}. @end deffn

@subsection Semantic Parsed System Databases

Another kind of system database is one created by the semantic tools, but saved in a special way. These databases are created with a script, or at some other convenient time. They can then be searched normally. They will appear on the same list of system databases as back ends that refer to object files indirectly.

A simple way to create one from Emacs is with this function:

@deffn Command semanticdb-create-system-database path &optional class @anchor{semanticdb-create-system-database} Create a system database starting at @var{path}. @var{path} should be a top level directory for a series of files containing declarations for @var{system} files. In @var{c}, this would be header filaes. @var{class} is the class for the database to create. Only child classes of symbol @dfn{semanticdb-project-database-system} are accepted. @end deffn

@node Idle Scheduling @chapter Idle Scheduling

The Semantic Idle Scheduler is a minor mode which performs semantic specific tasks in idle time. See @inforef{Idle Scheduler, , semantic-user.info}.

It performs the following tasks in order:

@enumerate @item Reprarse the current buffer if needed @item Reparse other buffers that need it @item Execute other scheduled semantic related operations. @end enumerate

Care is take in the idle scheduler to exit immediatly if user input is detected, improving editing performance.

The reason for grouping these tasks together is so that the automatic reparsing code executes before other idle services. This allows lexically broken buffers to be detected so that the other applications that follow do not accidentally reparse the buffer leaving unmatched syntax all over.

You can create new minor modes that are automatically scheduled by the semantic idle scheduler. Create the new minor mode with:

@defun define-semantic-idle-service name doc &rest forms @anchor{define-semantic-idle-service} Create a new idle services with @var{NAME}. @var{DOC} will be a documentation string describing @var{FORMS}. @var{FORMS} will be called during idle time after the current buffer's semantic tag information has been updated. This routines creates the following functions and variables: @end defun

@section User Input Handling

When writing an idle service, it is important for tasks that can take a long time to correctly exit upon user input.

You can test for user input in your idle handler with the following routines:

@defun semantic-throw-on-input from @anchor{semantic-throw-on-input} Exit with @dfn{throw} when in @dfn{semantic-exit-on-input} on user input. @var{FROM} is an indication of where this function is called from as a value to pass to @dfn{throw}. It is recommended to use the name of the function calling this one. @end defun

If you need to carefully extract from your function, you can wrap just a section of your function to exit on user input by wrapping it with this macro:

@defun semantic-exit-on-input symbol &rest forms @anchor{semantic-exit-on-input} Using @var{SYMBOL} as an argument to @dfn{throw}, execute @var{FORMS}. If @var{FORMS} includes a call to @code{semantic-thow-on-input}, then if a user presses any key during execution, this form macro will exit with the value passed to @dfn{semantic-throw-on-input}. If @var{FORMS} completes, then the return value is the same as @dfn{progn}. @end defun

Upon catching user input, you can try to detect if there was an exit from the return argument, and continue throwing with an additional call to @code{semantic-throw-on-input}.

@node Example Programs @chapter Programming Examples

*** NOTE *** These examples are for semantic 1.4. Using the below functions will generate compile time warnings with advice on what functions to use in semantic 2.0.

Here are some simple examples that use different aspects of the semantic library APIs. For fully functional example programs with lots of comments, see the file @file{semantic-examples.el}.

@heading Interactively querying for a token name

If you need a command that asks the user for a token name, you can get full range completion using the query functions @ref{Tag Completion}.

@example (interactive (list (semantic-read-symbol "Symbol: "))) @end example

@heading Finding a symbol in a buffer

If you have the name of a function or variable, and need to find its location in a buffer, you need a search function. There is a wide range of searches you can perform @ref{Searching Tag Tables}.

@example (semantic-find-nonterminal-by-name
"some-name"
(current-buffer)
t ;; look inside structures and classes for these symbols
nil) ;; do not look inside header files. @end example

@heading Finding a symbol in a project

If you have the name of a function or variable, and need to find its location somewhere in a project, you need to use the Semantic Database @inforef{semanticdb, , semanticdb}. There are many search functions similar to the ones found in @ref{Searching Tag Tables}.

The Semantic Database is interesting in that the return structure is not

@heading Locating a token in a buffer

If you have a nonterminal token, or a list of them, you may want to find their position in a buffer.

@example (semanticdb-find-nonterminal-by-name
"symbol"
nil ;; Defaults to the current project's database list.
t ;; Search inside types
nil ;; Do not search include files
nil ;; Only search files in the same mode (all C files)
t ;; When a token is found, make sure it is loaded in a buffer.
) @end example

Of interesting note above, semanticdb can find symbols in files that are not loaded into an Emacs buffer. These tokens do not have an associated overlay, and the function @dfn{semantic-token-buffer} will fail.

The last parameter's tells the search function to @dfn{find-file-noselect} any file in which a matching token was found. This will allow you to merge all the tokens into a completion list, or other flat list needed by most functions that use association lists.

If you do not ask semanticdb to load those files, you will need to explicitly request the database object (found in the @code{car} of each sublist) get the file loaded. It is useful to not auto find all files if you don't need to jump to that token.

@heading Converting a token into a human readable string.

A tag is a rather unpleasant Lisp structure when trying to decipher what is going on. As such, there is a wide range of functions available that can convert a token into a human readable, and colorful string @ref{Format Tag}.

If you program interfaces with lots of users, you will probably want to have your program define a configurable variable that will let users change the visible portion of your program.

@example (defcustom my-summary-function 'semantic-uml-prototype-nonterminal
"*Function to use when showing info about my tag."
:group 'my-program
:type semantic-format-tag-custom-list) @end example

Note the special type provided by Semantic.

Next, you can call this function to create a string.

@example (funcall my-summary-function tag
tag-parent
t ; use color
) @end example

In this case, @var{tag-parent} is an optional argument. In many cases, parent is not used by the outputting function. The parent must be a tag whose @code{semantic-tag-componenets} contains @var{tag}, or nil for top-level definitions. In particular, C++ needs the parent to correctly calculate the protection of each method.

@node Current Context @chapter Deriving the Current Context

This chapter deals with how to derive the current context, and also how a language maintainer can get the current context API to work with their language.

By default, the behavior will function in C like languages. This means languages with parenthetical blocks, and type dereferencing which uses a similar form.

@menu * Blocks:: * Local Variables:: Getting lists of local variables. * Derived Context:: What goes at a given location? * Context Analysis:: Analysis information about the local context. @end menu

@node Blocks @section Blocks and Navigation

Source code is typically built up of control structures, and blocks of context, or lexical scope. Semantic terms these lexical scopes as a ``context''. The following functions can be used to navigate contexts. Some of them are override functions. Language authors can override a subset of them to make them work for their language.

@defun semantic-up-context &optional point bounds-type @anchor{semantic-up-context} Move point up one context from @var{point}. Return non-@code{nil} if there are no more context levels. Overloaded functions using @code{up-context} take no parameters. @var{bounds-type} is a symbol representing a tag class to restrict movement to. If this is @code{nil}, @code{'function} is used. This will find the smallest tag of that class (function, variable, type, etc) and make sure non-@code{nil} is returned if you cannot go up past the bounds of that tag. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-beginning-of-context &optional point @anchor{semantic-beginning-of-context} Move @var{point} to the beginning of the current context. Return non-@code{nil} if there is no upper context. The default behavior uses @dfn{semantic-up-context}. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-end-of-context &optional point @anchor{semantic-end-of-context} Move @var{point} to the end of the current context. Return non-@code{nil} if there is no upper context. Be default, this uses @dfn{semantic-up-context}, and assumes parenthetical block delimiters. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

These next set of functions can be used to navigate across commands.

@defun semantic-end-of-command @anchor{semantic-end-of-command} Move to the end of the current command. Be default, uses @code{semantic-command-separation-character}. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-beginning-of-command @anchor{semantic-beginning-of-command} Move to the beginning of the current command. Be default, uses @code{semantic-command-separation-character}. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@node Local Variables @section Deriving local variables

Within a given context, or block of code, local variables are often defined. These functions can be used to retrieve lists of locally scoped variables.

@defun semantic-get-local-variables &optional point @anchor{semantic-get-local-variables} Get the local variables based on POINT's context. Local variables are returned in Semantic tag format. This can be overriden with @code{get-local-variables}. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-get-local-arguments &optional point @anchor{semantic-get-local-arguments} Get arguments (variables) from the current context at @var{point}. Parameters are available if the point is in a function or method. Return a list of tags unlinked from the originating buffer. Arguments are obtained by overriding @code{get-local-arguments}, or by the default function @dfn{semantic-get-local-arguments-default}. This, must return a list of tags, or a list of strings that will be converted to tags. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-get-all-local-variables &optional point @anchor{semantic-get-all-local-variables} Get all local variables for this context, and parent contexts. Local variables are returned in Semantic tag format. Be default, this gets local variables, and local arguments. Optional argument @var{point} is the location to start getting the variables from. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@node Derived Context @section Deriving the Current Context

While a context has already been used to describe blocks of code, other context include more local details, such as the symbol the cursor is on, or the fact we are assigning into some other variable.

These context deriving functions can be overridden to provide language specific behavior. By default, it assumes a C like language.

@defun semantic-ctxt-current-symbol &optional point @anchor{semantic-ctxt-current-symbol} Return the current symbol the cursor is on at @var{point} in a list. This will include a list of type/field names when applicable. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-ctxt-current-assignment &optional point @anchor{semantic-ctxt-current-assignment} Return the current assignment near the cursor at @var{point}. Return a list as per @dfn{semantic-ctxt-current-symbol}. Return @code{nil} if there is nothing relevant. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-ctxt-current-function &optional point @anchor{semantic-ctxt-current-function} Return the current function call the cursor is in at @var{point}. The function returned is the one accepting the arguments that the cursor is currently in. It will not return function symbol if the cursor is on the text representing that function. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-ctxt-current-argument &optional point @anchor{semantic-ctxt-current-argument} Return the index of the argument position the cursor is on at @var{point}. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-ctxt-current-thing @anchor{semantic-ctxt-current-thing} Calculate a thing identified by the current cursor position. Calls previously defined @code{semantic-ctxt-current-@dots{}} calls until something gets a match. See @dfn{semantic-ctxt-current-symbol}, @dfn{semantic-ctxt-current-function}, and @dfn{semantic-ctxt-current-assignment} for details on the return value. @end defun

@defun semantic-ctxt-current-class-list &optional point @anchor{semantic-ctxt-current-class-list} Return a list of tag classes that are allowed at @var{point}. If @var{point} is @code{nil}, the current buffer location is used. For example, in Emacs Lisp, the symbol after a ( is most likely a function. In a makefile, symbols after a : are rules, and symbols after a $( are variables. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@defun semantic-ctxt-scoped-types &optional point @anchor{semantic-ctxt-scoped-types} Return a list of type names currently in scope at @var{point}. The return value can be a mixed list of either strings (names of types that are in scope) or actual tags (type declared locally that may or may not have a name.) This function can be overloaded (see @dfn{define-mode-local-override} for details). @end defun

@node Context Analysis @section Analysis of the current context

The context parsing API is used in a context analysis library. This library provides high level routines for scanning through token databases to create lists of token associates. At it's core is a set of EIEIO classes defining a context. The context contains information about what was parsed at a given position, such as the strings there, and they type of assignment. The analysis library then searches the databases to determine the types and names available.

Two high level functions which can be run interactively are:

@deffn Command semantic-analyze-current-context &optional position @anchor{semantic-analyze-current-context} Analyze the current context at optional @var{position}. If called interactively, display interesting information about @var{position} in a separate buffer. Returns an object based on symbol @dfn{semantic-analyze-context}.

This function can be overriden with the symbol @code{analyze-context}. When overriding this function, your override will be called while cursor is at @var{position}. In addition, your function will not be called if a cached copy of the return object is found. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end deffn

@deffn Command semantic-analyze-possible-completions context @anchor{semantic-analyze-possible-completions} Return a list of semantic tags which are possible completions. @var{context} is either a position (such as point), or a precalculated context. Passing in a context is useful if the caller also needs to access parts of the analysis. Completions run through the following filters:

@itemize @item Elements currently in scope @item Constants currently in scope @item Elements match the @code{:prefix} in the @var{context}. @item Type of the completion matches the type of the context. @end itemize

Context type matching can identify the following:

@itemize @item No specific type @item Assignment into a variable of some type. @item Argument to a function with type constraints. @end itemize

When called interactively, displays the list of possible completions in a buffer. This function can be overloaded (see @dfn{define-mode-local-override} for details). @end deffn

@menu * Analysis Overview:: A description of how the analyzer works. * Analysis Objects:: What is in the analysis object. * Completion Overview:: How completions are calculated. @end menu

@node Analysis Overview @subsection Analysis Overview

The semantic analysis function @dfn{semantic-analye-current-context} creates an Analysis Object. See @ref{Analysis Objects}. This object contains many useful pieces of information needed to do any other kind of intelligent action on the local context.

If you call this function interactively, it will popup a buffer with a summary of the return value. This is useful when debugging.

@example +--------+ +----------------+ +----------------------------+ | Buffer |---| Context Parser |---| Local Context Synax Result | +--------+ +----------------+ +----------------------------+
| | +--------+ +-----------+ | | Parser |---| Tag Table |------------+ | +--------+ +-----------+ | |
V V +-------------+ +-------------------+ | Semantic DB |-------------------->| Semantic Analyzer | +-------------+ +-------------------+
|
V
+-----------------+
| Analysis Object |
+-----------------+ @end example

@node Analysis Objects @subsection Analysis Objects

@defindex analyze

@c First, delete text between this comment and END OF ANALYZE OBJECT text. @c @c Execute the below line with C-x C-e @c

@c (progn (require 'semantic-analyze) (eieiodoc-class semantic-analyze-context "analyze"))

@c END OF ANALZE OBJECT SECTION

@node Completion Overview @subsection Completion Overview

@node App Debugger @chapter Application level Data structure debugger

The data structures that Semantic provides can be complex, and figuring out why some level of application API performs incorrectly can be difficult.

The semantic Application debugger provides a way to look inside the various data structures of Semantic in a structures and complete way to help identify what a problem may be.

@section App Debugger Entry Points

There are a few basic functions that enter into the debugger:

@deffn Command semantic-adebug-bovinate @anchor{semantic-adebug-bovinate} The same as @dfn{bovinate}. Display the results in a debug buffer. @end deffn

@deffn Command semantic-adebug-searchdb regex @anchor{semantic-adebug-searchdb} Search the semanticdb for @var{regex} for the current buffer. Display the results as a debug list. @end deffn

@deffn Command semantic-adebug-analyze @anchor{semantic-adebug-analyze} Perform @dfn{semantic-analyze-current-context}. Display the results as a debug list. @end deffn

@section adebug-mode

The semantic debugger mode provides a simple user facing UI for looking into the data structures.

@deffn Command semantic-adebug-mode @anchor{semantic-adebug-mode} Major-mode for the Analyzer debugger.

Keybindings: @table @kbd @item spc semantic-adebug-expand-or-contract @item n semantic-adebug-next-expando @item p semantic-adebug-prev-expando @item n semantic-adebug-next @item p semantic-adebug-prev @item <mouse-2> semantic-adebug-expand-or-contract-mouse @end table

@end deffn

@section Create new debugger entry commands

Creating a new Application debugger entry point is easy. First, get a datastructure you need to analyze.

The first function to call is:

@defun semantic-adebug-new-buffer name @anchor{semantic-adebug-new-buffer} Create a new adebug buffer with @var{name}. @end defun

Next, you need to find the correct function for inserting your datastructure. All adebug insertion functions are of the form @code{semantic-adebug-insert-THING}, where @var{thing} is whatever you object is. Use Emacs help to pick something out.

@node GNU Free Documentation License @appendix GNU Free Documentation License

@include fdl.texi

@node Index @unnumbered Index @printindex cp

@iftex @contents @summarycontents @end iftex

@bye

@c Following comments are for the benefit of ispell.