[Mumps2Py:] [78] creating an m2p file now creates an idx file for the input file, which is now used instead of parsing through the entire input file every time .

[subversion@xxxxxxxxxxxxx]

Revision: 78
Author:   pgallot
Date:     2008-02-01 20:20:02 +0000 (Fri, 01 Feb 2008)

Log Message:
-----------
creating an m2p file now creates an idx file for the input file, which is now used instead of parsing through the entire input file every time.

Modified Paths:
--------------
    trunk/mumps2py/mumps2py_config.py
    trunk/mumps2py/mumps_module.py


Modified: trunk/mumps2py/mumps2py_config.py
===================================================================
--- trunk/mumps2py/mumps2py_config.py	2008-02-01 20:15:49 UTC (rev 77)
+++ trunk/mumps2py/mumps2py_config.py	2008-02-01 20:20:02 UTC (rev 78)
@@ -17,9 +17,9 @@
 
 """mumps2py_config contains all code for reading and writing any configuration
 settings related to mumps2py's translation of Mumps code to Python."""
-import ConfigParser
+import ConfigParser, os.path
 
-def create_m2p_file(m2p_fname, input_fname, output_dir = None):
+def create_m2p_file(m2p_fname, input_fname, output_dir = "./out"):
     """ creates a new m2p file."""
     config = ConfigParser.ConfigParser()
     config.set("DEFAULT", "in_filter_id", 0)
@@ -29,14 +29,32 @@
     config.add_section("filter_in")
     config.add_section("filter_out")
     config.set("input", "file", input_fname)
-    if output_dir:
-        config.set("output", "dir", output_dir)
-    else:
-        config.set("output", "dir", "./out")
+    config.set("output", "dir", output_dir)
+
+    idx_fname = create_idx_file(m2p_fname, input_fname, output_dir)
+    config.set("input", "index_file", idx_fname)
+    
     m2p_file = open(m2p_fname, 'w+')
     config.write(m2p_file)
     m2p_file.close()
 
+def create_idx_file(m2p_fname, input_fname, output_dir):
+    """creates an index file for a given m2p file"""
+    from mumps_module import parse_for_routines
+
+    idx_fname = os.path.splitext(m2p_fname)[0] + ".idx"
+    idx = ConfigParser.ConfigParser()
+    for num, routine in enumerate(parse_for_routines(input_fname, output_dir)):
+        section = "routine_%06d" % (num + 1)
+        idx.add_section(section)
+        idx.set(section, "name", routine.mod_name)
+        idx.set(section, "start", routine.start)
+        idx.set(section, "end", routine.end)
+    idx_file = open(idx_fname, 'w+')
+    idx.write(idx_file)
+    idx_file.close()
+    return idx_fname
+
 def add_m2p_filter_out_rule(m2p_fname, rule):
     """ adds a new exclusion filter rule to an m2p filter"""
     config = ConfigParser.ConfigParser()
@@ -71,10 +89,12 @@
     """extracts the info from an m2p file into a dictionary"""
     config = ConfigParser.ConfigParser()
     config.read(m2p_fname)
+    idx_fname = config.get("input", "index_file")
     info = {}
 
     info['inputfile'] = config.get("input", "file")
     info['outputdir'] = config.get("output", "dir")
+    info['index'] = []
     info['filter_out'] = []
     info['filter_in'] = []
     
@@ -86,4 +106,11 @@
         for rule in config.options("filter_in"):
             info['filter_in'].append(config.get("filter_in", rule))
 
+    idx = ConfigParser.ConfigParser()
+    idx.read(idx_fname)
+    for routine in idx.sections():
+        info['index'].append((idx.get(routine, "name"),
+                              idx.getint(routine, "start"),
+                              idx.getint(routine, "end")))
+    info['index'].sort()
     return info

Modified: trunk/mumps2py/mumps_module.py
===================================================================
--- trunk/mumps2py/mumps_module.py	2008-02-01 20:15:49 UTC (rev 77)
+++ trunk/mumps2py/mumps_module.py	2008-02-01 20:20:02 UTC (rev 78)
@@ -96,7 +96,7 @@
         self.filter_in_patterns.append ( re.compile(in_re_expr))
 
     
-def parse_cache_routines(inputfile, outputdir, routine_filter = None):
+def parse_cache_routines(inputfile, outputdir):
     """parseForModules returns a list of ModuleInfo objects from a file
 in Cache Format, assuming that each Mumps module begins
 with a line containing modulename^INT^"""
@@ -115,20 +115,19 @@
         if modpattern.search(line):
             tmp = modsplit.split(line, 1)
             if oldmodname:
-                if not routine_filter or routine_filter.accept(oldmodname):
-                    modules.append( \
-                        ModuleInfo(inputfile, outputdir, \
-                                   oldmodname, lastmodline + 1, lineno - 1))
-                if __debug__:
-                    print oldmodname, lastmodline + 1, lineno - 1, "\t", \
-                          lineno - lastmodline
+                modules.append( \
+                    ModuleInfo(inputfile, outputdir,
+                               oldmodname, lastmodline + 1, lineno - 1))
+            if __debug__:
+                print oldmodname, lastmodline + 1, lineno - 1, "\t", \
+                      lineno - lastmodline
             oldmodname = tmp[0]
             lastmodline = lineno
 
     fileinput.close()
     return modules                  
 
-def parse_astextroutines(inputfile, outputdir, routine_filter = None):
+def parse_astextroutines(inputfile, outputdir):
     """returns a list of ModuleInfo object from a file which has
 each routine separated by a blank line and the routine name given
 on its own line before the routine body."""
@@ -154,13 +153,11 @@
         if modmatch and emptypattern.match(last_line):
             tmp = modmatch.group("name")
             if oldmodname:
-                if not routine_filter or routine_filter.accept(oldmodname):
-                    modules.append( \
-                        ModuleInfo(inputfile, outputdir, \
-                                   oldmodname, lastmodline + 1, lineno - 1))
-                if __debug__:
-                    print oldmodname, lastmodline + 1, lineno - 1, "\t", \
-                          lineno - lastmodline
+                modules.append( ModuleInfo(inputfile, outputdir, oldmodname, 
+                                           lastmodline + 1, lineno - 1))
+            if __debug__:
+                print oldmodname, lastmodline + 1, lineno - 1, "\t", \
+                      lineno - lastmodline
             oldmodname = tmp
             lastmodline = lineno
         
@@ -182,16 +179,15 @@
     for filter_re in info['filter_in']:
         routine_filter.add_in_filter(filter_re)
 
-    mfile = open(inputfile)
-    headerline = mfile.readline()
-    mfile.close()
+    modules=[]
+    for routine, start, end in info['index']:
+        if routine_filter.accept(routine):
+            modules.append( ModuleInfo(inputfile, outputdir,
+                                       routine, start, end))
+            if __debug__:
+                print routine, start, end, "\t", end - start +1
 
-    if re.match(r"VistA FOIA Routines", headerline):
-        return parse_astextroutines(inputfile, outputdir, routine_filter)
-    elif re.search(r"CACHE FORMAT\^~Format=Cache.S~", headerline):
-        return parse_cache_routines(inputfile, outputdir, routine_filter)
-    else:
-        return None
+    return modules        
             
 def parse_for_routines(inputfile, outputdir):
     """returns a list of ModuleInfo objects"""