From d19f02083604b3f999f9ccb486e8e6c9bd7de4d8 Mon Sep 17 00:00:00 2001
From: Han-Wen Nienhuys <hanwen@lilypond.org>
Date: Tue, 13 Feb 2007 18:01:35 +0100
Subject: [PATCH] cache inventory, saving on lookup time.

---
 darcs2git.py | 55 +++++++++++++++++++++++++++++++++++--------------------
 1 file changed, 35 insertions(+), 20 deletions(-)

diff --git a/darcs2git.py b/darcs2git.py
index 72533c9..4285aa1 100644
--- a/darcs2git.py
+++ b/darcs2git.py
@@ -10,7 +10,9 @@
 # - time zones
 # - file modes
 # - use binary search to find from-patch ica. conflict.
-#
+# - use checkpointing to optimize speed?
+# - use get --partial ?
+
 
 import glob
 import os  
@@ -105,17 +107,17 @@ test end result.""")
     if not args:
         p.print_help ()
         sys.exit (2)
-        
+
+    options.basename = os.path.basename (args[0]).replace ('.darcs', '')
     if not options.target_git_repo:
-        p = args[0]
-        p = os.path.abspath (p)
-        options.target_git_repo = os.path.basename (p).replace ('.darcs', '')
-        options.target_git_repo += '.git' 
-       
+        options.target_git_repo = options.basename + '.git'
+        
     if options.debug:
         global log_file
         name = options.target_git_repo.replace ('.git', '.log')
-
+        if name == options.target_git_repo:
+            name += '.log'
+            
         progress ("Shell log to %s" % name)
         log_file = open (name, 'w')
         
@@ -163,15 +165,16 @@ class DarcsConversionRepo:
 
         self._current_number = -1
         self._is_valid = -1
-
-    def is_valid (self):
+        self._inventory_dict = None
+        
+    def is_contiguous (self):
         if not self._is_valid:
             return False
 
         darcs_dir = self.dir + '/_darcs'
         inv = self.inventory ()
         for p in self.patches[:self._current_number + 1]:
-            if p.short_id () not in inv:
+            if not self.has_patch (p):
                 return False
 
         if self._current_number + 1 != len (inv.split ('\n[')):
@@ -180,13 +183,9 @@ class DarcsConversionRepo:
         return True
     
     def has_patch (self, p):
-        id = p.attributes['hash']
-        darcs_dir = self.dir + '/_darcs'
-        f = darcs_dir + '/patches/' + id
-        if not os.path.exists (f):
-            return False
+        assert self._is_valid
         
-        return p.short_id () in self.inventory ()
+        return self.inventory_dict ().has_key (p.short_id ())
     
     def pristine_tree (self):
         return self.dir + '/_darcs/pristine'
@@ -198,7 +197,10 @@ class DarcsConversionRepo:
         dir = self.dir
         progress ('Rewinding %d patches' % count)
         system ('cd %(dir)s && echo ay|darcs obliterate --ignore-times --last %(count)d' % locals ())
-
+        d = self.inventory_dict () 
+        for p in self.patches[self._current_number - count:self._current_number+1]:
+            del d[p.short_id ()] 
+        
     def clean (self):
         system ('rm -rf %s' % self.dir)
 
@@ -212,15 +214,17 @@ class DarcsConversionRepo:
         dir = self.dir
 
         progress ('Pull patch %d' % patch.number)
-        self._current_number = patch.number
         system ('cd %(dir)s && darcs pull --ignore-times --quiet --all --match "hash %(id)s" %(source_repo)s ' % locals ())
 
+        self._current_number = patch.number
+
     def create_fresh (self):
         dir = self.dir
         system ('rm -rf %(dir)s && mkdir %(dir)s && darcs init --repo  %(dir)s'
                 % locals ())
         self._is_valid = True
         self._current_number = -1
+        self._inventory_dict = None
         
     def inventory (self):
         darcs_dir = self.dir + '/_darcs'
@@ -229,6 +233,17 @@ class DarcsConversionRepo:
             i += open (f).read ()
         return i
 
+    def inventory_dict (self):
+        if type (self._inventory_dict) == type ({}):
+            return self._inventory_dict
+
+        self._inventory_dict = {}
+
+        def note_patch (m):
+            self._inventory_dict[m.group (1)] = 1
+
+        re.sub (r'\n([^*\n])+\*\*([0-9]+)', note_patch, self.inventory ())
+            
     def start_at (self, num):
         progress ('Go to patch %d' % num)
 
@@ -260,7 +275,7 @@ class DarcsConversionRepo:
         It might be quicker and/or more correct to wind/rewind the
         repo with pull and unpull."""
 
-        valid = self.is_valid ()
+        valid = self.is_contiguous ()
         where = ''
         if valid:
             where  = 'at %d' % self._current_number
-- 
2.11.4.GIT