Telemetry: add a metadata layer between page set and .wpr.

tools/telemetry/telemetry/page_set_archive_info.py

Index: tools/telemetry/telemetry/page_set_archive_info.py
diff --git a/tools/telemetry/telemetry/page_set_archive_info.py b/tools/telemetry/telemetry/page_set_archive_info.py
new file mode 100644
index 0000000000000000000000000000000000000000..e77ed2fb17853344ec84594f56c0db3df2b4c239
--- /dev/null
+++ b/tools/telemetry/telemetry/page_set_archive_info.py
@@ -0,0 +1,118 @@
+# Copyright (c) 2013 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+import collections
+import json
+import logging
+import os
+import re
+
+class PageSetArchiveInfo(object):
+  def __init__(self, archive_data_file_path, page_set_file_path, data):
+    self._archive_data_file_path = archive_data_file_path
+    self._archive_data_file_dir = os.path.dirname(archive_data_file_path)
+    # Back pointer to the page set file.
+    self._page_set_file_path = page_set_file_path
+
+    # Map from the relative path (as it appears in the metadata file) of the
+    # .wpr file to a list of urls it supports.
+    self._wpr_file_to_urls = collections.OrderedDict(data['archives'])
+
+    # Map from the page url to a relative path (as it appears in the metadata
+    # file) of the .wpr file.
+    self._url_to_wpr_file = dict()
+    # Find out the wpr file names for each page.
+    for wpr_file in data['archives']:
+      page_urls = data['archives'][wpr_file]
+      for url in page_urls:
+        self._url_to_wpr_file[url] = wpr_file
+
+  @classmethod
+  def FromFile(cls, file_path, page_set_file_path):
+    with open(file_path, 'r') as f:
+      data = json.load(f)
+      return cls(file_path, page_set_file_path, data)
+
+  def WprFileForPage(self, page):
+    return self._url_to_wpr_file.get(page.url, None)
+
+  def WprFilePathForPage(self, page):
+    wpr_file = self.WprFileForPage(page)
+    if wpr_file:
+      return self._WprFileNameToPath(wpr_file)
+    return None
+
+  def AddNewRecording(self, pages):
+    (target_wpr_file, target_wpr_file_path) = self._NextWprFileName()
+    for page in pages:
+      self._SetWprFileForPage(page, target_wpr_file)
+    return target_wpr_file_path
+
+  def DeleteAbandonedWprFiles(self):
+    # Update the metadata so that the abandoned wpr files don't have empty url
+    # arrays.
+    abandoned_wpr_files = self.AbandonedWprFiles()
+    for wpr_file in abandoned_wpr_files:
+      del self._wpr_file_to_urls[wpr_file]
+      # Don't fail if we're unable to delete some of the files.
+      wpr_file_path = self._WprFileNameToPath(wpr_file)
+      try:
+        os.remove(wpr_file_path)
+      except Exception:
+        logging.warning('Failed to delete file: %s' % wpr_file_path)
+
+  def WriteToFile(self):
+    """Writes the metadata into the file passed as constructor parameter."""
+    metadata = dict()
+    metadata['description'] = (
+        'Describes the Web Page Replay archives for a page set. Don\'t edit by '
+        'hand! Use record_wpr for updating.')
+    # Pointer from the metadata to the page set .json file.
+    metadata['page_set'] = os.path.relpath(self._page_set_file_path,
+                                           self._archive_data_file_dir)
+    metadata['archives'] = self._wpr_file_to_urls.copy()
+    # Don't write data for abandones archives.
+    abandoned_wpr_files = self.AbandonedWprFiles()
+    for wpr_file in abandoned_wpr_files:
+      del metadata['archives'][wpr_file]
+
+    with open(self._archive_data_file_path, 'w') as f:
+      json.dump(metadata, f, indent=4)
+      f.flush()
+
+  def AbandonedWprFiles(self):
+    abandoned_wpr_files = []
+    for wpr_file, urls in self._wpr_file_to_urls.iteritems():
+      if not urls:
+        abandoned_wpr_files.append(wpr_file)
+    return abandoned_wpr_files
+
+  def _WprFileNameToPath(self, wpr_file):
+    return os.path.abspath(os.path.join(self._archive_data_file_dir, wpr_file))
+
+  def _NextWprFileName(self):
+    """Creates a new file name for a wpr archive file."""
+    # The names are of the format "some_thing_number.wpr". Read the numbers.
+    highest_number = -1
+    base = None
+    for wpr_file in self._wpr_file_to_urls:
+      match = re.match(r'(?P<BASE>.*)_(?P<NUMBER>[0-9]+)\.wpr', wpr_file)
+      if not match:
+        raise Exception('Illegal wpr file name ' + wpr_file)
+      highest_number = max(int(match.groupdict()['NUMBER']), highest_number)
+      if base and match.groupdict()['BASE'] != base:
+        raise Exception('Illegal wpr file name ' + wpr_file +
+                        ', doesn\'t begin with ' + base)
+      base = match.groupdict()['BASE']
+    new_filename = '%s_%03d.wpr' % (base, highest_number + 1)
+    return new_filename, self._WprFileNameToPath(new_filename)
+
+  def _SetWprFileForPage(self, page, wpr_file):
+    """For modifying the metadata when we're going to record a new archive."""
+    old_wpr_file = self.WprFileForPage(page)
+    if old_wpr_file:
+      self._wpr_file_to_urls[old_wpr_file].remove(page.url)
+    self._url_to_wpr_file[page.url] = wpr_file
+    if wpr_file not in self._wpr_file_to_urls:
+      self._wpr_file_to_urls[wpr_file] = []
+    self._wpr_file_to_urls[wpr_file].append(page.url)