Merge pull request #354 from nexB/cache-and-stream-scan-results

Cache and stream scan results

Author: pombredanne

setup.py

@@ -78,7 +78,7 @@ def read(*names, **kwargs):
         # caching
         'zc.lockfile >= 1.0.0, < 2.0.0',
         'yg.lockfile >= 2.0.0, < 3.0.0',
-        'diskcache >= 1.7.0, < 1.8.0',
+        'diskcache >= 2.0.0, < 3.0.0',
 
         # textcode
         'Beautifulsoup >= 3.2.0, <4.0.0',
@@ -99,6 +99,7 @@ def read(*names, **kwargs):
         'jinja2 >= 2.7.0, < 3.0.0',
         'MarkupSafe >= 0.23',
         'colorama',
+        'simplejson',
 
         # packagedcode
         'requests >= 2.7.0, < 3.0.0',

src/scancode/__init__.py

@@ -1,5 +1,5 @@
 #
-# Copyright (c) 2015 nexB Inc. and others. All rights reserved.
+# Copyright (c) 2016 nexB Inc. and others. All rights reserved.
 # http://nexb.com and https://github.com/nexB/scancode-toolkit/
 # The ScanCode software is licensed under the Apache License version 2.0.
 # Data generated with ScanCode require an acknowledgment.
@@ -23,3 +23,21 @@
 #  Visit https://github.com/nexB/scancode-toolkit/ for support and download.
 
 __version__ = '2.0.0rc2'
+
+from os.path import dirname
+from os.path import abspath
+from os.path import getsize
+from os.path import getmtime
+from os.path import join
+from os.path import exists
+
+from commoncode import fileutils
+
+scan_src_dir = abspath(dirname(__file__))
+src_dir = dirname(scan_src_dir)
+root_dir = dirname(src_dir)
+cache_dir = join(root_dir, '.cache')
+scans_cache_dir = join(cache_dir, 'scan_results_caches')
+
+if not exists(scans_cache_dir):
+    fileutils.create_dir(scans_cache_dir)

src/scancode/api.py

@@ -142,11 +142,11 @@ def get_licenses(location, min_score=0):
             result['matched_rule']['licenses'] = match.rule.licenses
             # TODO: add debug details such as matcher
             # result['matched_rule']['matcher'] = match.matcher
-            
+
             yield result
 
 
-def get_file_infos(location):
+def get_file_infos(location, as_list=True):
     """
     Return a list of dictionaries of informations collected from the file or
     directory at location.
@@ -177,7 +177,10 @@ def get_file_infos(location):
     infos['is_media'] = is_file and T.is_media or None
     infos['is_source'] = is_file and T.is_source or None
     infos['is_script'] = is_file and T.is_script or None
-    return [infos]
+    if as_list:
+        return [infos]
+    else:
+        return infos
 
 
 def get_package_infos(location):

src/scancode/cache.py

@@ -0,0 +1,162 @@
+#
+# Copyright (c) 2016 nexB Inc. and others. All rights reserved.
+# http://nexb.com and https://github.com/nexB/scancode-toolkit/
+# The ScanCode software is licensed under the Apache License version 2.0.
+# Data generated with ScanCode require an acknowledgment.
+# ScanCode is a trademark of nexB Inc.
+#
+# You may not use this software except in compliance with the License.
+# You may obtain a copy of the License at: http://apache.org/licenses/LICENSE-2.0
+# Unless required by applicable law or agreed to in writing, software distributed
+# under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
+# CONDITIONS OF ANY KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations under the License.
+#
+# When you publish or redistribute any data created with ScanCode or any ScanCode
+# derivative work, you must accompany this data with the following acknowledgment:
+#
+#  Generated with ScanCode and provided on an "AS IS" BASIS, WITHOUT WARRANTIES
+#  OR CONDITIONS OF ANY KIND, either express or implied. No content created from
+#  ScanCode should be considered or used as legal advice. Consult an Attorney
+#  for any legal advice.
+#  ScanCode is a free software code scanning tool from nexB Inc. and others.
+#  Visit https://github.com/nexB/scancode-toolkit/ for support and download.
+
+from __future__ import absolute_import, print_function
+
+from collections import OrderedDict
+import os
+import sys
+
+from commoncode import fileutils
+from commoncode import timeutils
+
+from scancode import scans_cache_dir
+
+"""
+Caching scans on disk: A cache of all the scan results.
+
+Each scan results for a file or directory is cached on disk.
+
+The approach is to use to cache:
+ - the results of a scan, excluding file infos keyed by the hash of a scanned file
+ - the file infos, keyed by the path of a scanned file
+
+Once a scan is completed, we iterate the caches to output the scan results using this
+procedure: iterate the cached file infos and for each lookup the scan details in the
+cached scan results. This iteration is driving the final streaming of results to the
+output format (e.g. JSON).
+
+Finally once a scan is completed the cache is destroyed to free up disk space.
+"""
+
+# Tracing flags
+TRACE = False
+
+def logger_debug(*args):
+    pass
+
+if TRACE:
+    import logging
+
+    logger = logging.getLogger(__name__)
+    # logging.basicConfig(level=logging.DEBUG, stream=sys.stdout)
+    logging.basicConfig(stream=sys.stdout)
+    logger.setLevel(logging.DEBUG)
+
+    def logger_debug(*args):
+        return logger.debug(' '.join(isinstance(a, basestring) and a or repr(a) for a in args))
+
+
+class ScanCache(object):
+    """
+    A file-based cache for scan results.
+    This is NOT thread-safe, but is multi-process safe.
+    """
+    def __init__(self, cache_dir):
+        fileutils.create_dir(cache_dir)
+
+        # create a unique temp directory in cache_dir
+        self.cache_base_dir = fileutils.get_temp_dir(cache_dir, prefix=timeutils.time2tstamp() + '-')
+
+        # and subdirs for infos and scans caches
+        self.cache_infos_dir = os.path.join(self.cache_base_dir, 'infos')
+        fileutils.create_dir(self.cache_infos_dir)
+        self.cache_scans_dir = os.path.join(self.cache_base_dir, 'scans')
+        fileutils.create_dir(self.cache_scans_dir)
+
+        # workaround for https://github.com/grantjenks/python-diskcache/issues/32
+        from diskcache import Disk
+        class DiskWithNoHighPickleProtocol(Disk):
+            "Subclass of diskcache.Disk that always use the lowest pickle protocol."
+            def __init__(self, directory, size_threshold, pickle_protocol):
+                super(DiskWithNoHighPickleProtocol, self).__init__(directory, size_threshold, pickle_protocol)
+                self._protocol = 0
+
+        # and finially cache instances
+        from diskcache import Cache
+        self.infos = Cache(self.cache_infos_dir, disk=DiskWithNoHighPickleProtocol)
+        self.scans = Cache(self.cache_scans_dir, disk=DiskWithNoHighPickleProtocol)
+
+    def scan_key(self, path, file_infos):
+        """
+        Return a scan cache key for a path and file_infos.
+        """
+        sha1 = file_infos['sha1']
+        # we may eventually store directories, in which case we use the path as a key
+        return sha1 or path
+
+    def put_infos(self, path, file_infos):
+        """
+        Put file_infos for path in the cache and return True if the file referenced
+        in file_infos has already been scanned or False otherwise.
+        """
+        self.infos.set(path, file_infos)
+        has_cached_details = self.scan_key(path, file_infos) in self.scans
+        if TRACE:
+            logger_debug('put_infos:', 'path:', path, 'has_cached_details:', has_cached_details, 'file_infos:', file_infos, '\n')
+            logger_debug('put_infos:', 'cached_infos:', self.infos[path], '\n')
+
+        return has_cached_details
+
+    def put_scan(self, path, file_infos, scan_result):
+        """
+        Put scan_result in the cache. Also put  file_infos in the cache if needed.
+        """
+        scan_key = self.scan_key(path, file_infos)
+        self.scans.add(scan_key, scan_result)
+        if TRACE:
+            logger_debug('put_scan:', 'scan_key:', scan_key, 'file_infos:', file_infos, 'scan_result:', scan_result, '\n')
+            logger_debug('put_scan:', 'cached_infos:', self.infos[path], '\n')
+            logger_debug('put_scan:', 'scan_key:', scan_key, 'cached_scan:', self.scans[scan_key], '\n')
+
+    def iterate(self, with_infos=True):
+        """
+        Return an iterator of scan data for all cached scans e.g. the whole cache.
+        """
+        for path in self.infos:
+            file_infos = self.infos[path]
+            scan_result = OrderedDict(path=path)
+            if with_infos:
+                # infos is always collected but only returnedd if asked:
+                # we flatten these as direct attributes of a file object
+                scan_result.update(file_infos.items())
+
+            scan_key = self.scan_key(path, file_infos)
+            scan_details = self.scans[scan_key]
+            scan_result.update(scan_details)
+            if TRACE:
+                logger_debug('iterate:', 'scan_details:', scan_details, 'for path:', path, 'scan_key:', scan_key, '\n')
+            yield scan_result
+
+    def clear(self, *args):
+        """
+        Purge the cache by deleting the corresponding cached data files.
+        """
+        self.infos.close()
+        self.scans.close()
+        fileutils.delete(self.cache_base_dir)
+
+
+def get_scans_cache():
+    return ScanCache(cache_dir=scans_cache_dir)