fix/parser storage stability

src/org/rascalmpl/library/IO.rsc

@@ -18,6 +18,7 @@ The following input/output functions are defined:
 module IO
 
 import Exception;
+
 extend analysis::diff::edits::FileSystemChanges;
 
 @synopsis{All functions in this module that have a charset parameter use this as default.}
@@ -209,7 +210,6 @@ exists(|std:///IO.rsc|);
 @javaClass{org.rascalmpl.library.Prelude}
 public java bool exists(loc file);
 
-
 @synopsis{Find a named file in a list of locations.}
 @examples{
 ```rascal-shell

src/org/rascalmpl/library/lang/rascal/grammar/definition/Modules.rsc

@@ -18,6 +18,8 @@ import Set;
 import IO;
 import util::Monitor;
 
+private loc here = |std:///lang/rascal/syntax/grammar/definition/Modules|;
+
 @memo
 @synopsis{Converts internal module representation of Rascal interpreter to single grammar definition}
 public Grammar modules2grammar(str main, map[str name, tuple[set[str] imports, set[str] extends, set[SyntaxDefinition] defs] \mod] mods) {
@@ -51,24 +53,30 @@ public Grammar fuse(GrammarDefinition def) {
   done = {};
   deps = dependencies(def);
 
+  if (!def.modules[def.main]?) {
+    jobWarning("the main module is unavailable <def.main> (ignored)", here);
+  }
+
   while (todo != {}) {
     <nm,todo> = takeOneFrom(todo);
     done += nm; 
-    if(def.modules[nm]?){
-        \mod = def.modules[nm];
-        result = (compose(result, \mod.grammar) | compose(it, def.modules[i].grammar) | i <- deps[nm], def.modules[i]?);
-        todo += (\mod.extends - done);
-    }
-    else {
-      warning("Fuse algorithm misses module definition for dependency <nm>", |unknown:///|);
+
+    \mod = def.modules[nm];
+
+    for (str i <- deps[nm], !def.modules[i]?) {
+      jobWarning("<nm> imports or extends the unavailable module <i> (ignored)", here);
     }
+
+    result = (compose(result, \mod.grammar) | compose(it, def.modules[i].grammar) | i <- deps[nm], def.modules[i]?);
+    todo += (\mod.extends - done);
   }
 
   return result;
 }
 
 
 
+
 public GrammarModule module2grammar(Module \mod) {
   <nm, imps, exts> = getModuleMetaInf(\mod);
   return \module(nm, imps, exts, syntax2grammar(collect(\mod)));

src/org/rascalmpl/library/lang/rascal/grammar/storage/ModuleParserStorage.rsc

@@ -102,13 +102,26 @@ int main(list[str] args) {
     storeParsersForModules(pcfg);
 }
 ```
+
+Or, you could use the ((PathConfig)) parameter feature of `main` functions. The Rascal maven plugin will
+make sure to pass the proper ((PathConfig)) parameter to your main function:
+```rascal
+module YourMainModule
+
+import util::Reflective;
+import lang::rascal::grammar::storage::ModuleParserStorage;
+
+int main(PathConfig pcfg = pathConfig()) {
+    storeParsersForModules(pcfg);
+}
+```
 }
 void storeParsersForModules(PathConfig pcfg) {
-    storeParsersForModules({*find(src, "rsc") | src <- pcfg.srcs, bprintln("Crawling <src>")}, pcfg);
+    storeParsersForModules({*find(src, "rsc", exclude={*pcfg.ignores}) | src <- pcfg.srcs, bprintln("Crawling <src>")}, pcfg);
 }
 
 void storeParsersForModules(set[loc] moduleFiles, PathConfig pcfg) {
-    storeParsersForModules({parseModule(m) | m <- moduleFiles, bprintln("Loading <m>")}, pcfg);
+    storeParsersForModules({parseModule(m) | m <- moduleFiles, m notin pcfg.ignores, bprintln("Loading <m>")}, pcfg);
 }
 
 void storeParsersForModules(set[Module] modules, PathConfig pcfg) {

src/org/rascalmpl/library/util/FileSystem.rsc

@@ -7,35 +7,94 @@
 }
 module util::FileSystem
 
+import Exception;
 import IO;
+import util::Monitor;
 
+@synopsis{Model of a file system with its (nested) files and directories}
 data FileSystem 
   = directory(loc l, set[FileSystem] children)
   | file(loc l)
   ;
 
-FileSystem crawl(loc l) = isDirectory(l) ? directory(l, {crawl(e) | e <- l.ls}) : file(l);
+@synopsis{Extract a compositional ((FileSystem)) model starting from a given directory location.}
+@description{
+* Using `exclude` you can avoid going into certain directories or filter specific files from the result.  
+* With `checkExist=true` the `l` parameter is checked to exist before the file system is crawled and a PathNotFound exception is thrown if not.
+}  
+FileSystem crawl(loc l, set[loc] exclude= {}, bool checkExist=true) throws PathNotFound
+  = isDirectory(l) ? directory(l, {crawl(e, exclude=exclude, checkExist=false) | e <- l.ls, l notin exclude}) : file(l)
+  when checkExist ==> throwNotExist(l)
+  ;
 
-@synopsis{Recursively lists locations of all files from the supplied directory.
-  If input is a file, its location is returned instead.}
-set[loc] files(loc l) = isDirectory(l) ? { *files(e) | e <- l.ls } : {l};
+@synopsis{Recursively lists locations of all files from the supplied directory.}
+@description{
+* If input `l` is a file, its location is returned instead.
+* Using `exclude` you can avoid going into certain directories or filter specific files from the result.  
+* With `checkExist=true` the `l` parameter is checked to exist before the file system is crawled and a PathNotFound exception is thrown if not.
+}
+set[loc] files(loc l, set[loc] exclude={}, bool checkExist=true) throws PathNotFound
+  = isDirectory(l) ? { *files(e, exclude=exclude, checkExist=false) | e <- l.ls, e notin exclude} : {l} 
+  when checkExist ==> throwNotExist(l);
 
-@synopsis{Recursively lists locations of all files that satisfy the filter criterion `filt`.
-  For a file to be included, `filt` must return `true` for it.}
-set[loc] find(loc f, bool (loc) filt) 
+@synopsis{Recursively lists locations of all files that satisfy the filter criterion `filt`.}
+@description{
+* For a file to be included, `filt` must return `true` for it. All directories are traversed though, regardless of `filt`.
+* Using `exclude` you can avoid going into certain directories or filter specific files from the result.  
+* With `checkExist=true` the `f` parameter is checked to exist before the file system is crawled and a PathNotFound exception is thrown if not.
+}
+set[loc] find(loc f, bool (loc) filt, set[loc] exclude = {}, bool checkExist=true) throws PathNotFound
   = isDirectory(f) 
-      ? {*find(c, filt) | c <- f.ls} + (filt(f) ? {f} : { }) 
+      ? {*find(c, filt, exclude=exclude, checkExist=false) | c <- f.ls, c notin exclude} + ((filt(f) && f notin exclude) ? {f} : { }) 
       : (filt(f) ? {f} : { })
+      when checkExist ==> throwNotExist(f)
       ;
 
 @synopsis{Recursively lists locations of all files that end in `ext`.}
-set[loc] find(loc f, str ext) = find(f, bool (loc l) { return l.extension == ext; });
+@description{
+* For a file to be included, it's extension must equal `ext`. All directories are traversed though, regardless of their extension.
+* Using `exclude` you can avoid going into certain directories or filter specific files from the result.  
+* With `checkExist=true` the `f` parameter is checked to exist before the file system is crawled and a PathNotFound exception is thrown if not.
+}
+set[loc] find(loc f, str ext, set[loc] exclude={}, bool checkExist=true) throws PathNotFound
+  = find(f, bool (loc l) { return l.extension == ext; }, exclude=exclude, checkExist=checkExist);
 
 @synopsis{Lists all files recursively ignored files and directories starting with a dot.}
-set[loc] visibleFiles(loc l) {
-  if (/^\./ := l.file) 
+set[loc] visibleFiles(loc l, bool checkExist=true) throws PathNotFound {
+  if (checkExist) {
+    throwNotExist(l);
+  }
+  if (/^\./ := l.file) {
     return {};
-  if (isDirectory(l)) 
+  }
+  if (isDirectory(l)) {
     return {*visibleFiles(f) | f <- l.ls}; 
+  }
   return {l};
 }
+
+@synopsis{Always returns true, but shows a ((util::Monitor::jobWarning)) if `file` does not exist.}
+@benefits{
+* This can be used practically in comprehensions that process file locations.
+* Use it to fail more transparantly, but still in case of erroneous file configuration (paths)
+}
+bool warnNotExist(loc file) {
+  if (!exists(file)) {
+    jobWarning("<file> does not exist.", |std:///util/FileSystem|);
+  }
+  return true;
+}
+
+@synopsis{Always returns true, except when throwing FileNotFound if `file` does not exist}
+@benefits{
+* This can be used practically in comprehensions that process file locations;
+* Use it to fail faster and harder in case of erroneous  file configuration (paths)
+}
+bool throwNotExist(loc file) throws PathNotFound {
+  if (!exists(file)) {
+    throw PathNotFound(file);
+  }
+  return true;
+}
+
+