aboutsummaryrefslogtreecommitdiff
path: root/libbuild2/parser.cxx
AgeCommit message (Collapse)AuthorFilesLines
2020-11-17Generalize dot escaping in target name rulesBoris Kolpackov1-1/+1
Now triple dot and escape sequence can appear almost anywhere in the target name (see target::split_name() for details).
2020-09-24Give hints for common causes of "no rule to update ..." errorBoris Kolpackov1-13/+14
2020-08-12Add int64 and int64s variable typesBoris Kolpackov1-0/+2
2020-07-14Recognize `build2` as special module name in addition to `build`Boris Kolpackov1-2/+2
This is for consistency with version constraints in manifest.
2020-07-13Fix version check in using directiveBoris Kolpackov1-8/+9
2020-07-13Reserve backtick (`) and bit-or (|) in eval context for future useBoris Kolpackov1-0/+8
Specifically, they are reserved for future support of arithmetic evaluation contexts and evaluation pipelines, respectively.
2020-07-12Rename rule-adhoc-* to adhoc-rule-*Boris Kolpackov1-2/+2
2020-07-09Add support for ad hoc importationBoris Kolpackov1-3/+8
2020-07-06Adjust variable block applicability in dependency chainsBoris Kolpackov1-106/+122
Before the block used to apply to the set of prerequisites before the last `:`. This turned out to be counterintuitive and not very useful since prerequisite-specific variables are a lot less common than target specific. And it doesn't fit with ad hoc recipes. The new rule is if the chain ends with `:`, then the block applies to the last set of prerequisites. Otherwise, it applies to the last set of targets. For example: ./: exe{test}: cxx{main} { test = true # Applies to the exe{test} target. } ./: exe{test}: libue{test}: { bin.whole = false # Applies to the libue{test} prerequisite. } This is actually consistent with both non-chain and non-block cases. Consider: exe{test}: cxx{main} { test = true } exe{test}: libue{test}: { bin.whole = false } exe{test}: libue{test}: bin.whole = false The only exception we now have in this overall approach of "if the dependency declaration ends with a colon, then what follows is for a prerequisite" is for the first semicolon: exe{test}: { test = true } exe{test}: test = true But that's probably intuitive enough since there cannot be a prerequisite without a target.
2020-06-18Delay checking for ad hoc C++ recipe support until matchBoris Kolpackov1-9/+1
2020-06-15Diagnose building of module or ad hoc C++ recipe using static build systemBoris Kolpackov1-2/+6
2020-06-11Diagnose use of ad hoc C++ recipes with bootstrap build systemBoris Kolpackov1-1/+5
2020-06-10Add ad hoc recipe if-else, switch tests (and fix bug)Boris Kolpackov1-4/+10
2020-06-10Handle special variable names when spelled as $(<char>) rather than $<char>Boris Kolpackov1-2/+41
2020-06-10Fix bug in subscript of NULL valuesBoris Kolpackov1-1/+3
2020-06-09Factor ad hoc C++ and Buildscript rules into separate filesBoris Kolpackov1-1/+4
2020-06-05Add depdb buildscript builtinKaren Arutyunov1-1/+1
2020-06-05Add ability to split ad hoc C++ recipe into global and local fragmentsBoris Kolpackov1-6/+34
Specifically, now we can write: {{ c++ 1 -- #include <map> -- recipe apply (action, target&) const override { ... } }}
2020-06-05Add ability to specify ad hoc recipe actionsBoris Kolpackov1-8/+122
We are reusing the buildspec syntax for that.
2020-06-04Properly handle diag directive in build script parserKaren Arutyunov1-1/+2
2020-06-03Allow process path values and targets as buildscript program namesKaren Arutyunov1-2/+2
Also deduce the recipe name.
2020-06-03Add versioning for ad hoc C++ recipesBoris Kolpackov1-2/+21
This will allow us to deal with backward-incompatible changes to cxx_rule interface and semantics.
2020-06-03Factor implementation-specific ad hoc recipe parsing to adhoc_*_ruleBoris Kolpackov1-68/+40
2020-05-29Move low-verbosity command name from adhoc_script_rule to scriptBoris Kolpackov1-3/+3
2020-05-29Add support for is-else, switch in ad hoc recipesBoris Kolpackov1-130/+262
2020-05-27Add support for value subscript after expansionsBoris Kolpackov1-27/+118
Value subscript is only recognized in evaluation contexts (due to ambiguity with wildcard patterns; consider: $x[123].txt) and should be unseparated from the previous token. For example: x = ($y[1]) x = (($f ? $y : $z)[1]) x = ($identity($y)[$z])
2020-05-27Initial support for ad hoc recipes (still work in progress)Boris Kolpackov1-140/+404
2020-05-27Improve ternary diagnostics with info for wildcard pattern escapeBoris Kolpackov1-2/+13
2020-05-01Fix outstanding issue with directive vs assignment differentiationBoris Kolpackov1-6/+14
Specifically, now the following does the right thing: print +foo
2020-04-30Rename target::member to target::adhoc_memberBoris Kolpackov1-2/+2
2020-04-27Rework tool importation along with cli moduleBoris Kolpackov1-55/+107
Specifically, now config.<tool> (like config.cli) is handled by the import machinery (it is like a shorter alias for config.import.<tool>.<tool>.exe that we already had). And the cli module now uses that instead of custom logic. This also adds support for uniform tool metadata extraction that is handled by the import machinery. As a result, a tool that follows the "build2 way" can be imported with metadata by the buildfile and/or corresponding module without any tool-specific code or brittleness associated with parsing --version or similar outputs. See the cli tool/module for details. Finally, two new flavors of the import directive are now supported: import! triggers immediate importation skipping any rule-specific logic while import? is optional import (analogous to using?). Note that optional import is always immediate. There is also the import-specific metadata attribute which can be specified for these two import flavors in order to trigger metadata importation. For example: import? [metadata] cli = cli%exe{cli} if ($cli != [null]) info "cli version $($cli:cli.version)"
2020-04-27Don't switch projects when switching scopes during bootstrapBoris Kolpackov1-5/+13
2020-04-27Fix assertion failure when name extension pattern is usedKaren Arutyunov1-0/+4
2020-04-08Allow configuration variables in unnamed projectsBoris Kolpackov1-9/+27
While generally a bad idea, there are valid situations where this may happen, such as a standalone build of the tests subproject in test-installed.
2020-03-31Handle duplicate config directives for same variableBoris Kolpackov1-1/+18
2020-03-31Switch to project variable visibility by defaultBoris Kolpackov1-2/+3
2020-03-27Fix bug in previous commitBoris Kolpackov1-2/+2
2020-03-27Implement project configuration reporting, similar to build system modulesBoris Kolpackov1-85/+186
2020-03-26Make buildfile parser reset'ableBoris Kolpackov1-12/+32
Note that the testscript parser (which derives from the buildfile parser) is (still) not reset'able (this functionality is currently not needed so why complicate things).
2020-03-25Enforce config directives only appearing in project's root.buildBoris Kolpackov1-6/+7
2020-03-20Initial implementation of config directive for project-specific configurationBoris Kolpackov1-1/+135
2020-03-20Lexer support for default value assignment (?=)Boris Kolpackov1-0/+10
Note: not yet supported in the parser.
2020-03-18Cleanup variable type/visibility/overridability logicBoris Kolpackov1-40/+54
2020-03-18Get rid of old and unused logicBoris Kolpackov1-30/+5
2020-03-17Rename all find*(variable) to lookup*(variable)Boris Kolpackov1-2/+2
Now we consistently use term "lookup" for variable value lookup. At some point we should also rename type lookup to binding and get rid of all the lookup_type aliases.
2020-02-07Drop copyright notice from source codeKaren Arutyunov1-1/+0
2020-01-28Use scope::var_pool()Boris Kolpackov1-3/+3
2020-01-27Improve module loading APIBoris Kolpackov1-1/+1
2020-01-27Improve process run_*() APIBoris Kolpackov1-9/+6
2019-11-15Test and document wildcard character escapingBoris Kolpackov1-1/+1
Also document the new bracket expression ([...]) wildcard support.