summaryrefslogtreecommitdiffstats
path: root/vhdlparser
Commit message (Collapse)AuthorAgeFilesLines
* Allow building doxygen as a subprojectAnonymous Maarten2020-09-161-4/+4
|
* Vhdl improvements (ALIAS, translation) (#7813)powARman2020-08-032-1/+6
| | | | | | | | | | | | | * Support VHDL alias constructs. * Translate class to "Design Unit" for VHDL. * Fix compile error * Add new function trDesignUnitDocumentation() to translator. Adapt english and german translation to use the new function. Co-authored-by: Andreas Regel <andreas.regel@newayselectronics.com>
* remove vhdlparser/MakefileCosta Shulyupin2020-06-282-54/+15
| | | | because it is generated from vhdlparser/CMakeLists.txt
* Regression: further fixes to make the documentation build againDimitri van Heesch2020-06-102-1579/+1214
|
* issue #7778 Build fails with javacc 5.0 (#7779)albert-github2020-06-101-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * issue #7778 Build fails with javacc 5.0 At the moment doxygen works with javaCC version 7.0.5 but no test is done for the version. The minimum requirement has been set to 7.0.5 javacc does not have an option to retrieve the version though when giving `--version` a bit more information is given, this information is filtered. (an issue has been submitted to javacc to obtain the version information: https://github.com/javacc/javacc/issues/172). * issue #7778 Build fails with javacc 5.0 Version detection did not work for `5.0` (only x.y.z was handled, not x.y) * issue #7778 Build fails with javacc 5.0 Based on a review - `--version` is not a valid option, but present already for future use but - removed double ` if(JAVACC_EXECUTABLE)` - Needed `"${JAVACC_TEMP_VERSION}"` as otherwise when `javacc` is found but cannot be executed a message like: ``` CMake Error at cmake/FindJavacc.cmake:11 (string): string sub-command REGEX, mode MATCH needs at least 5 arguments total to command. Call Stack (most recent call first): vhdlparser/CMakeLists.txt:1 (find_package) ``` would appear - reformulated some status strings * issue #7778 Build fails with javacc 5.0 the `.` was not correctly escaped. * issue #7778 Build fails with javacc 5.0 Another one 7.0........ match, going for the suggested match in the cmake CheckCXXCompilerFlag * issue #7778 Build fails with javacc 5.0 Accidently left debug statement. * issue #7778 Build fails with javacc 5.0 Problem has been fixed upstream for version 7.0.7 and up, not using `--version` but `-version` as all command line arguments are starting with a single '-'. Also it is just the bare version, so small extra test so it will work with all versions.
* missing instance-reworkhake2020-06-031-1/+1
|
* Added generated filesDimitri van Heesch2020-05-122-723/+735
|
* fixed if-generate-bug 7721hake2020-05-111-15/+15
|
* Migrated some code in pre.l to use STL containersDimitri van Heesch2020-03-282-490/+246
|
* Fix 6342: Applying provided patchJonas Fingerling2020-03-264-2509/+2885
|
* Enabled stricter compiler warnings and fixed all new warningsDimitri van Heesch2020-03-081-6/+3
|
* Warnings in case of a VHDL erroralbert-github2020-02-161-5/+5
| | | | | | | | | | | | | | | | | | | | When having a vhdl construct like: ``` entity _H_ is ``` we get the, a bit unclear, warnings like: ``` Lexical error at: 16:8. Encountered: _ after: . ``` with version 1.8.17 we got the warning: ``` .../vhdl.vhd:16: warning: Lexical error, Encountered: '_' after: '' ``` not 100% clear either but at least clear which file is involved and what the meaning of the '16' is. Also the message didn't conform the doxygen style / place / handling of warnings / errors anymore (didn't go to the WARN_LOGFILE anymore). (The message is correct as a vhdl identifier cannot start or end with an underscore. Between the used javaCC versions (now 7.05) and the previously used version 6.xx apparently a small in the error handler prototypes has been introduces, and thus not finding the doxygen version anymore.
* Fix build error after 'make clean' due to creation of unpatched JavaCC.hDimitri van Heesch2020-02-121-1/+1
|
* issue #7411: "warning: return type of member is not documented" for static ↵Dimitri van Heesch2020-02-093-2166/+2151
| | | | void and virtual void functions
* vhdl-fixed-bug-7432hake2020-02-041-36/+35
|
* Regenerate checked in code with JavaCC 7.0.5Dimitri van Heesch2020-01-2517-8011/+6510
|
* fixed some parser bugs,make parser ready for javacc 7.0.5hake2020-01-242-65/+131
|
* Make VHDL parser reentrantDimitri van Heesch2020-01-038-1860/+1808
|
* Merge branch 'master' into spellingDimitri van Heesch2019-12-2317-4240/+6669
|\
| * Merge branch 'master' of github.com:doxygen/doxygenDimitri van Heesch2019-12-121-1/+1
| |\
| | * Merge pull request #7440 from albert-github/feature/bug_JAVACC_FLAGSDimitri van Heesch2019-12-121-1/+1
| | |\ | | | | | | | | Adding JAVACC_FLAGS for javacc compilation
| | | * Adding JAVACC_FLAGS for javacc compilationalbert-github2019-12-121-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | During transition to cmake build the, essential for debugging, flag possibility was not added. (Added analogous to LEX_FLAGS and YACC_Flags). Available debug options (according to old Makefile): `-debug_parser` `-debug_token_manager` `-debug_lookahead`
| * | | Minor fixesDimitri van Heesch2019-12-122-87/+121
| | | |
| * | | Merge branch 'feature/issue_7436' of ↵Dimitri van Heesch2019-12-121-6/+6
| |\ \ \ | | |/ / | |/| | | | | | https://github.com/albert-github/doxygen into albert-github-feature/issue_7436
| | * | issue #7436 Incorrect handling of block comments in VHDLalbert-github2019-12-111-6/+6
| | |/ | | | | | | | | | | | | | | | The search for `/*` or /*!` ended at the last `*/` in a file and thus skipping other intermediate block end and new starts. Also the intermediate code was lost see as comment. The filter pattern used was incorrect and should have been `<"/*" (~["*"])* "*" ("*" | ~["*","/"] (~["*"])* "*")* "/">` (thanks to https://javacc.org/contrib/C.jj). Here also the space plus one or more `*` at the end beginning of the line are still incorporated as well as multiple `*` before the colosing `*/` this is also filtered.
| * | Revert to locally patched version 6.2 of JavaCC due to more regressions :-(Dimitri van Heesch2019-12-1214-609/+668
| | |
| * | Revert to use of javacc-6.1.3 due to regressionDimitri van Heesch2019-12-1215-4724/+7059
| |/
| * Merge pull request #7334 from albert-github/feature/issue_7313_vhdl_subtypeDimitri van Heesch2019-12-101-0/+3
| |\ | | | | | | issue #7313 VHDL attribute 'subtype is not supported and breaks parser
| | * issue #7313 VHDL attribute 'subtype is not supported and breaks parseralbert-github2019-10-181-0/+3
| | | | | | | | | | | | Add the `'subtype` construct for variables etc.
| * | Changed std::unique_ptr<Entry> to std::shared_ptr<Entry> at avoid use after ↵Dimitri van Heesch2019-12-083-6/+4
| | | | | | | | | | | | free issues
* | | spelling: occursJosh Soref2019-11-121-1/+1
| | |
* | | spelling: implementationJosh Soref2019-11-121-1/+1
| | |
* | | spelling: followingJosh Soref2019-11-122-2/+2
|/ /
* | Update of generated VhdlParser files.Dimitri van Heesch2019-11-062-5/+5
| |
* | Spelling corrections for vhdl directoryalbert-github2019-11-041-1/+1
| | | | | | | | | | | | | | | | Spelling corrections as found by codespell and in #561. `assume_guarentee` is not a vhdl keyword, `assume_guarantee` is. A number of spelling errors are in generated code (through javacc) and have been reported upstrean (https://github.com/javacc/javacc/pull/118) Corection of `guarded_signal_specificatio` to `guarded_signal_specification` is not possible as `guarded_signal_specification` exists as well.
* | Fix double delete in VHDL error handler.Dimitri van Heesch2019-11-031-2/+0
| |
* | Update generated VHDL code to JavaCC 7.0.5Dimitri van Heesch2019-11-0316-4850/+2514
| |
* | Update generated code to javacc 6.2Dimitri van Heesch2019-10-2814-617/+672
| |
* | Adding possibility to automatically build vhdlparser from vhdlparser.jj filealbert-github2019-10-181-0/+10
|/ | | | | - possibility to find the javacc program / script - creating command to compile javaparser.jj with javacc and place files in the right directory when javacc is present.
* Discrepancy between vhdl input and generated sourcesalbert-github2019-10-171-2/+2
| | | | | | The generated vhdl code files (.cc and .h) and didn't match the vhdl input source (vhdlpardser.jj). When regenerating with javacc the vhdl code file they didn't compile. vhdlparser.jj has been corrected.
* Use smartpointers to manage the lifetime of Entry objectsDimitri van Heesch2019-10-073-23/+23
|
* Fix #6906Ton van den Heuvel2019-05-061-9/+3
| | | | | | Generate configvalues.h only as a dependency for the VHDL parser. Fixes a problem in combination with Ninja v1.9.0 because of duplicate rules for generating configvalues.h.
* Improved const correctness and added support for inline namespacesDimitri van Heesch2019-04-211-1/+1
|
* Bug 691689 - Line numbers for examplesDimitri van Heesch2018-08-192-6/+6
|
* Merge pull request #686 from albert-github/feature/bug_vhdl_error_messageDimitri van Heesch2018-05-012-6/+23
|\ | | | | Making VHDL error messages more doxygen like
| * Making VHDL error messages more doxygen likealbert-github2018-03-272-6/+23
| |
* | Misc. typosluz.paz2018-04-253-3/+3
| | | | | | | | Found via `codespell`
* | Adding debug options to vhdl parser generatoralbert-github2018-03-262-5/+45
|/ | | | | | | | JavaCC contains by default options to get more information about the parser paths used (analogous to the -d option of flex). - Makefile - adding option (JAVACC_FLAGS) for easier use of JavaCC options - adding possibility to reset generated files to their git versions. - vhdlstring.h adding functions required by debug options (debug_token_manager)
* Restore Makefile in vhdlparser dirDimitri van Heesch2017-10-281-397/+22
|
* Marks JS as freely licensedzachwick2017-08-041-22/+397
| | | | | | | | | | | | | | | Marking the resulting JS in this way ensures that visitors to the resulting HTML documentation do not have to choose between running non-free JS and experiencing the HTML documentation as it was intended. The JS was already freely licensed, so this change just ensures that the appropriate labelling occurs so that GNU LibreJS [0] parses the JS correctly as freely licensed. [0] https://www.gnu.org/software/librejs/ Signed-off-by: zachwick <zach@zachwick.com>
generated by cgit v0.12 at 2024-11-01 13:06:32 (GMT)