Fix various minor linearization script issues

- The last-timestamp-encountered variable wasn’t being used properly. Rewrite code to properly allow for new blockchain files to be written when split by month.
- Properly set a blockchain file’s access and modify times.
- Add a “debug output” option to quiet certain output that might not always be desirable.
- Update the README.

3 files changed, 41 insertions, 18 deletions

diff --git a/contrib/linearize/README.md b/contrib/linearize/README.md
index adc9a559cc..0971e7816b 100644
--- a/contrib/linearize/README.md
+++ b/contrib/linearize/README.md
@@ -32,8 +32,11 @@ Required configuration file settings:
 * `output`: Output directory for linearized `blocks/blkNNNNN.dat` output.
 
 Optional config file setting for linearize-data:
-* `file_timestamp`: Set each file's last-modified time to that of the most
-recent block in that file.
+* `debug_output`: Some printouts may not always be desired. If true, such output
+will be printed.
+* `file_timestamp`: Set each file's last-accessed and last-modified times,
+respectively, to the current time and to the timestamp of the most recent block
+written to the script's blockchain.
 * `genesis`: The hash of the genesis block in the blockchain.
 * `input`: bitcoind blocks/ directory containing blkNNNNN.dat
 * `hashlist`: text file containing list of block hashes created by
@@ -41,6 +44,9 @@ linearize-hashes.py.
 * `max_out_sz`: Maximum size for files created by the `output_file` option.
 (Default: `1000*1000*1000 bytes`)
 * `netmagic`: Network magic number.
+* `out_of_order_cache_sz`: If out-of-order blocks are being read, the block can
+be written to a cache so that the blockchain doesn't have to be seeked again.
+This option specifies the cache size. (Default: `100*1000*1000 bytes`)
 * `rev_hash_bytes`: If true, the block hash list written by linearize-hashes.py
 will be byte-reversed when read by linearize-data.py. See the linearize-hashes
 entry for more information.
diff --git a/contrib/linearize/example-linearize.cfg b/contrib/linearize/example-linearize.cfg
index cccdd79213..db53dc0ef6 100644
--- a/contrib/linearize/example-linearize.cfg
+++ b/contrib/linearize/example-linearize.cfg
@@ -1,4 +1,3 @@
-
 # bitcoind RPC settings (linearize-hashes)
 rpcuser=someuser
 rpcpassword=somepassword
@@ -21,6 +20,9 @@ input=/home/example/.bitcoin/blocks
 #genesis=000000000933ea01ad0ee984209779baaec3ced90fa3f408719526f8d77f4943
 #input=/home/example/.bitcoin/testnet3/blocks
 
+# "output" option causes blockchain files to be written to the given location,
+# with "output_file" ignored. If not used, "output_file" is used instead.
+# output=/home/example/blockchain_directory
 output_file=/home/example/Downloads/bootstrap.dat
 hashlist=hashlist.txt
 
@@ -29,3 +31,12 @@ out_of_order_cache_sz = 100000000
 
 # Do we want the reverse the hash bytes coming from getblockhash?
 rev_hash_bytes = False
+
+# On a new month, do we want to set the access and modify times of the new
+# blockchain file?
+file_timestamp = 0
+# Do we want to split the blockchain files given a new month or specific height?
+split_timestamp = 0
+
+# Do we want debug printouts?
+debug_output = False
diff --git a/contrib/linearize/linearize-data.py b/contrib/linearize/linearize-data.py
index 3fdec134b8..afcec2b60a 100755
--- a/contrib/linearize/linearize-data.py
+++ b/contrib/linearize/linearize-data.py
@@ -134,7 +134,7 @@ class BlockDataCopier:
 		if not self.fileOutput and ((self.outsz + blockSizeOnDisk) > self.maxOutSz):
 			self.outF.close()
 			if self.setFileTime:
-				os.utime(outFname, (int(time.time()), highTS))
+				os.utime(self.outFname, (int(time.time()), self.highTS))
 			self.outF = None
 			self.outFname = None
 			self.outFn = self.outFn + 1
@@ -142,12 +142,12 @@ class BlockDataCopier:
 
 		(blkDate, blkTS) = get_blk_dt(blk_hdr)
 		if self.timestampSplit and (blkDate > self.lastDate):
-			print("New month " + blkDate.strftime("%Y-%m") + " @ " + hash_str)
-			lastDate = blkDate
-			if outF:
-				outF.close()
-				if setFileTime:
-					os.utime(outFname, (int(time.time()), highTS))
+			print("New month " + blkDate.strftime("%Y-%m") + " @ " + self.hash_str)
+			self.lastDate = blkDate
+			if self.outF:
+				self.outF.close()
+				if self.setFileTime:
+					os.utime(self.outFname, (int(time.time()), self.highTS))
 				self.outF = None
 				self.outFname = None
 				self.outFn = self.outFn + 1
@@ -155,11 +155,11 @@ class BlockDataCopier:
 
 		if not self.outF:
 			if self.fileOutput:
-				outFname = self.settings['output_file']
+				self.outFname = self.settings['output_file']
 			else:
-				outFname = os.path.join(self.settings['output'], "blk%05d.dat" % self.outFn)
-			print("Output file " + outFname)
-			self.outF = open(outFname, "wb")
+				self.outFname = os.path.join(self.settings['output'], "blk%05d.dat" % self.outFn)
+			print("Output file " + self.outFname)
+			self.outF = open(self.outFname, "wb")
 
 		self.outF.write(inhdr)
 		self.outF.write(blk_hdr)
@@ -223,13 +223,16 @@ class BlockDataCopier:
 			blk_hdr = self.inF.read(80)
 			inExtent = BlockExtent(self.inFn, self.inF.tell(), inhdr, blk_hdr, inLen)
 
-			hash_str = calc_hash_str(blk_hdr)
-			if not hash_str in blkmap:
-				print("Skipping unknown block " + hash_str)
+			self.hash_str = calc_hash_str(blk_hdr)
+			if not self.hash_str in blkmap:
+				# Because blocks can be written to files out-of-order as of 0.10, the script
+				# may encounter blocks it doesn't know about. Treat as debug output.
+				if settings['debug_output'] == 'true':
+					print("Skipping unknown block " + self.hash_str)
 				self.inF.seek(inLen, os.SEEK_CUR)
 				continue
 
-			blkHeight = self.blkmap[hash_str]
+			blkHeight = self.blkmap[self.hash_str]
 			self.blkCountIn += 1
 
 			if self.blkCountOut == blkHeight:
@@ -295,12 +298,15 @@ if __name__ == '__main__':
 		settings['max_out_sz'] = 1000 * 1000 * 1000
 	if 'out_of_order_cache_sz' not in settings:
 		settings['out_of_order_cache_sz'] = 100 * 1000 * 1000
+	if 'debug_output' not in settings:
+		settings['debug_output'] = 'false'
 
 	settings['max_out_sz'] = int(settings['max_out_sz'])
 	settings['split_timestamp'] = int(settings['split_timestamp'])
 	settings['file_timestamp'] = int(settings['file_timestamp'])
 	settings['netmagic'] = unhexlify(settings['netmagic'].encode('utf-8'))
 	settings['out_of_order_cache_sz'] = int(settings['out_of_order_cache_sz'])
+	settings['debug_output'] = settings['debug_output'].lower()
 
 	if 'output_file' not in settings and 'output' not in settings:
 		print("Missing output file / directory")