Compare sha1 hashes when restoring

git-svn-id: http://svn.savannah.nongnu.org/svn/rdiff-backup/trunk@666 2b77aa54-bcbc-44c9-a7ec-4f6cf2b41109

6 files changed, 45 insertions, 10 deletions

diff --git a/rdiff-backup/CHANGELOG b/rdiff-backup/CHANGELOG
index a0e1fa5..534ecbe 100644
--- a/rdiff-backup/CHANGELOG
+++ b/rdiff-backup/CHANGELOG
@@ -2,7 +2,11 @@ New in v1.1.1 (????/??/??)
 --------------------------
 
 rdiff-backup now writes SHA1 sums into its mirror_metadata file for
-all regular files.
+all regular files, and checks them when restoring.
+
+Added two new comparison modes: full file (using the --compare-full or
+--compare-full-at-time) or by hash (--compare-hash and
+--compare-hash-at-time).
 
 Applied Alec Berryman's patch to update the no-compression regexp.
 
diff --git a/rdiff-backup/rdiff-backup.1 b/rdiff-backup/rdiff-backup.1
index beb1941..6f94955 100644
--- a/rdiff-backup/rdiff-backup.1
+++ b/rdiff-backup/rdiff-backup.1
@@ -12,7 +12,9 @@ rdiff-backup \- local/remote mirror and incremental backup
 .BI "| --remove-older-than " time_interval
 .BI "| --list-at-time " time
 .BI "| --list-changed-since " time
-.B "| --list-increment-sizes "}
+.B "| --list-increment-sizes "
+.B "| --verify"
+.BI "| --verify-at-time " time }
 .BI [[[ user@ ] host2.foo ]:: destination_directory ]
 
 .B rdiff-backup --calculate-average
@@ -427,6 +429,15 @@ section for more information.
 Specify verbosity level (0 is totally silent, 3 is the default, and 9
 is noisiest).  This determines how much is written to the log file.
 .TP
+.B --verify
+This is short for
+.BI --verify-at-time " now"
+.TP
+.BI --verify-at-time " now"
+Check all the data in the repository at the given time by computing
+the SHA1 hash of all the regular files and comparing them with the
+hashes stored in the metadata file.
+.TP
 .B "-V, --version"
 Print the current version and exit
 
diff --git a/rdiff-backup/rdiff_backup/Main.py b/rdiff-backup/rdiff_backup/Main.py
index 62ab840..04bb4e0 100644
--- a/rdiff-backup/rdiff_backup/Main.py
+++ b/rdiff-backup/rdiff_backup/Main.py
@@ -728,7 +728,7 @@ def Compare(compare_type, src_rp, dest_rp, compare_time = None):
 	return_val = compare_func(src_rp, mirror_rp, inc_rp, compare_time)
 
 def Verify(dest_rp, verify_time = None):
-	"""Check the hashs of the regular files against mirror_metadata"""
+	"""Check the hashes of the regular files against mirror_metadata"""
 	global return_val
 	dest_rp = require_root_set(dest_rp, 1)
 	if not verify_time:
diff --git a/rdiff-backup/rdiff_backup/Rdiff.py b/rdiff-backup/rdiff_backup/Rdiff.py
index f183141..3ed589d 100644
--- a/rdiff-backup/rdiff_backup/Rdiff.py
+++ b/rdiff-backup/rdiff_backup/Rdiff.py
@@ -87,6 +87,9 @@ def patch_local(rp_basis, rp_delta, outrp = None, delta_compressed = None):
 	file (librsync may need to seek around in it).  If outrp is None,
 	patch rp_basis instead.
 
+	The return value is the close value of the delta, so it can be
+	used to produce hashes.
+
 	"""
 	assert rp_basis.conn is Globals.local_connection
 	if delta_compressed: deltafile = rp_delta.open("rb", 1)
diff --git a/rdiff-backup/rdiff_backup/compare.py b/rdiff-backup/rdiff_backup/compare.py
index 2395395..0679b22 100644
--- a/rdiff-backup/rdiff_backup/compare.py
+++ b/rdiff-backup/rdiff_backup/compare.py
@@ -195,7 +195,7 @@ class DataSide(backup.SourceStruct):
 
 	def compare_hash(cls, repo_iter):
 		"""Like above, but also compare sha1 sums of any regular files"""
-		def hashs_changed(src_rp, mir_rorp):
+		def hashes_changed(src_rp, mir_rorp):
 			"""Return 0 if their data hashes same, 1 otherwise"""
 			if not mir_rorp.has_sha1():
 				log.Log("Warning: Metadata file has no digest for %s, "
@@ -208,7 +208,7 @@ class DataSide(backup.SourceStruct):
 
 		src_iter = cls.get_source_select()
 		for src_rp, mir_rorp in rorpiter.Collate2Iters(src_iter, repo_iter):
-			report = get_basic_report(src_rp, mir_rorp, hashs_changed)
+			report = get_basic_report(src_rp, mir_rorp, hashes_changed)
 			if report: yield report
 			else: log_success(src_rp, mir_rorp)
 
diff --git a/rdiff-backup/rdiff_backup/restore.py b/rdiff-backup/rdiff_backup/restore.py
index 8f0a5a5..a97f839 100644
--- a/rdiff-backup/rdiff_backup/restore.py
+++ b/rdiff-backup/rdiff_backup/restore.py
@@ -22,7 +22,7 @@
 from __future__ import generators
 import tempfile, os, cStringIO
 import Globals, Time, Rdiff, Hardlink, rorpiter, selection, rpath, \
-	   log, static, robust, metadata, statistics, TempFile, eas_acls
+	   log, static, robust, metadata, statistics, TempFile, eas_acls, hash
 
 
 class RestoreError(Exception): pass
@@ -256,7 +256,8 @@ class MirrorStruct:
 			mir_rorp.flaglinked(Hardlink.get_link_index(mir_rorp))
 		elif mir_rorp.isreg():
 			expanded_index = cls.mirror_base.index + mir_rorp.index
-			mir_rorp.setfile(cls.rf_cache.get_fp(expanded_index))
+			file_fp = cls.rf_cache.get_fp(expanded_index)
+			mir_rorp.setfile(hash.FileWrapper(file_fp))
 		mir_rorp.set_attached_filetype('snapshot')
 		return mir_rorp
 
@@ -616,15 +617,31 @@ class PatchITRB(rorpiter.ITRBranch):
 		self.patch_to_temp(rp, diff_rorp, tf)
 		rpath.rename(tf, rp)
 
+	def check_hash(self, copy_report, diff_rorp):
+		"""Check the hash in the copy_report with hash in diff_rorp"""
+		if not diff_rorp.isreg(): return
+		if not diff_rorp.has_sha1():
+			log.Log("Hash for %s missing, cannot check" %
+					(diff_rorp.get_indexpath()), 2)
+		elif copy_report.sha1_digest == diff_rorp.get_sha1():
+			log.Log("Hash %s of %s verified" %
+					(diff_rorp.get_sha1(), diff_rorp.get_indexpath()), 6)
+		else:
+			log.Log("Warning: Hash %s of %s\ndoesn't match recorded hash %s!"
+					% (copy_report.sha1_digest, diff_rorp.get_indexpath(),
+					   diff_rorp.get_sha1()), 2)
+
 	def patch_to_temp(self, basis_rp, diff_rorp, new):
 		"""Patch basis_rp, writing output in new, which doesn't exist yet"""
 		if diff_rorp.isflaglinked():
 			Hardlink.link_rp(diff_rorp, new, self.basis_root_rp)
-		elif diff_rorp.get_attached_filetype() == 'snapshot':
-			rpath.copy(diff_rorp, new)
+			return
+		if diff_rorp.get_attached_filetype() == 'snapshot':
+			copy_report = rpath.copy(diff_rorp, new)
 		else:
 			assert diff_rorp.get_attached_filetype() == 'diff'
-			Rdiff.patch_local(basis_rp, diff_rorp, new)
+			copy_report = Rdiff.patch_local(basis_rp, diff_rorp, new)
+		self.check_hash(copy_report, diff_rorp)
 		if new.lstat(): rpath.copy_attribs(diff_rorp, new)
 
 	def start_process(self, index, diff_rorp):