X-Git-Url: https://arthur.barton.de/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=lib%2Fbup%2Fhashsplit.py;h=048347b2fdd312be787c7cfd3dbf7d6eb9820446;hb=de7ccab9ea2b7fa0937ca1aea5d88d97b06893cd;hp=1819294d1aa7d2029203316d78f9c300bc75b7c8;hpb=9133f733cdde36d7ecd627d339f90d87b7d2b0e6;p=bup.git

diff --git a/lib/bup/hashsplit.py b/lib/bup/hashsplit.py
index 1819294..048347b 100644
--- a/lib/bup/hashsplit.py
+++ b/lib/bup/hashsplit.py
@@ -8,6 +8,11 @@ MAX_PER_TREE = 256
 progress_callback = None
 fanout = 16
 
+GIT_MODE_FILE = 0100644
+GIT_MODE_TREE = 040000
+GIT_MODE_SYMLINK = 0120000
+assert(GIT_MODE_TREE != 40000)  # 0xxx should be treated as octal
+
 # The purpose of this type of buffer is to avoid copying on peek(), get(),
 # and eat().  We do copy the buffer contents on put(), but that should
 # be ok if we always only put() large amounts of data at a time.
@@ -36,6 +41,13 @@ class Buf:
         return len(self.data) - self.start
 
 
+def _fadvise_done(f, ofs, len):
+    assert(ofs >= 0)
+    assert(len >= 0)
+    if len > 0 and hasattr(f, 'fileno'):
+        _helpers.fadvise_done(f.fileno(), ofs, len)
+
+
 def readfile_iter(files, progress=None):
     for filenum,f in enumerate(files):
         ofs = 0
@@ -43,11 +55,12 @@ def readfile_iter(files, progress=None):
         while 1:
             if progress:
                 progress(filenum, len(b))
-            fadvise_done(f, max(0, ofs - 1024*1024))
             b = f.read(BLOB_READ_SIZE)
             ofs += len(b)
+            # Warning: ofs == 0 means 'done with the whole file'
+            # This will only happen here when the file is empty
+            _fadvise_done(f, 0, ofs)
             if not b:
-                fadvise_done(f, ofs)
                 break
             yield b
 
@@ -56,11 +69,13 @@ def _splitbuf(buf, basebits, fanbits):
     while 1:
         b = buf.peek(buf.used())
         (ofs, bits) = _helpers.splitbuf(b)
-        if ofs > BLOB_MAX:
-            ofs = BLOB_MAX
         if ofs:
+            if ofs > BLOB_MAX:
+                ofs = BLOB_MAX
+                level = 0
+            else:
+                level = (bits-basebits)//fanbits  # integer division
             buf.eat(ofs)
-            level = (bits-basebits)//fanbits  # integer division
             yield buffer(b, 0, ofs), level
         else:
             break
@@ -115,17 +130,20 @@ def split_to_blobs(makeblob, files, keep_boundaries, progress):
 
 def _make_shalist(l):
     ofs = 0
+    l = list(l)
+    total = sum(size for mode,sha,size, in l)
+    vlen = len('%x' % total)
     shalist = []
     for (mode, sha, size) in l:
-        shalist.append((mode, '%016x' % ofs, sha))
+        shalist.append((mode, '%0*x' % (vlen,ofs), sha))
         ofs += size
-    total = ofs
+    assert(ofs == total)
     return (shalist, total)
 
 
 def _squish(maketree, stacks, n):
     i = 0
-    while i<n or len(stacks[i]) > MAX_PER_TREE:
+    while i < n or len(stacks[i]) >= MAX_PER_TREE:
         while len(stacks) <= i+1:
             stacks.append([])
         if len(stacks[i]) == 1:
@@ -133,7 +151,7 @@ def _squish(maketree, stacks, n):
         elif stacks[i]:
             (shalist, size) = _make_shalist(stacks[i])
             tree = maketree(shalist)
-            stacks[i+1].append(('40000', tree, size))
+            stacks[i+1].append((GIT_MODE_TREE, tree, size))
         stacks[i] = []
         i += 1
 
@@ -145,29 +163,29 @@ def split_to_shalist(makeblob, maketree, files,
     if not fanout:
         shal = []
         for (sha,size,level) in sl:
-            shal.append(('100644', sha, size))
+            shal.append((GIT_MODE_FILE, sha, size))
         return _make_shalist(shal)[0]
     else:
         stacks = [[]]
         for (sha,size,level) in sl:
-            stacks[0].append(('100644', sha, size))
-            if level:
-                _squish(maketree, stacks, level)
+            stacks[0].append((GIT_MODE_FILE, sha, size))
+            _squish(maketree, stacks, level)
         #log('stacks: %r\n' % [len(i) for i in stacks])
         _squish(maketree, stacks, len(stacks)-1)
         #log('stacks: %r\n' % [len(i) for i in stacks])
         return _make_shalist(stacks[-1])[0]
 
 
-def split_to_blob_or_tree(makeblob, maketree, files, keep_boundaries):
+def split_to_blob_or_tree(makeblob, maketree, files,
+                          keep_boundaries, progress=None):
     shalist = list(split_to_shalist(makeblob, maketree,
-                                    files, keep_boundaries))
+                                    files, keep_boundaries, progress))
     if len(shalist) == 1:
         return (shalist[0][0], shalist[0][2])
     elif len(shalist) == 0:
-        return ('100644', makeblob(''))
+        return (GIT_MODE_FILE, makeblob(''))
     else:
-        return ('40000', maketree(shalist))
+        return (GIT_MODE_TREE, maketree(shalist))
 
 
 def open_noatime(name):
@@ -180,9 +198,3 @@ def open_noatime(name):
         except:
             pass
         raise
-
-
-def fadvise_done(f, ofs):
-    assert(ofs >= 0)
-    if ofs > 0 and hasattr(f, 'fileno'):
-        _helpers.fadvise_done(f.fileno(), ofs)