def Compare(base, compare, ops, root_path=None, out_path=None):
"""Compares a series of scrapes using a series of operators.
Args:
base: (browser, version) tuple of version to consider the baseline
compare: (browser, version) tuple of version to compare to
ops: list of operators plus operator arguments
root_path: root of the scrapes
out_path: place to put any output from the operators
Returns:
None
@TODO(jhaas): this method will likely change, to provide a robust and
well-defined way of chaining operators, applying operators conditionally,
and full-featured scripting of the operator chain. There also needs
to be better definition of the output; right now it's to stdout and
a log.txt file, with operator-dependent images saved for error output
"""
if root_path is None:
# default save path is "scrapes" off the current root
root_path = os.path.join(os.path.split(__file__)[0], "Scrapes")
if out_path is None:
out_path = os.path.join(os.path.split(__file__)[0], "Compares")
if type(base) in types.StringTypes: base = (base, None)
if type(compare) in types.StringTypes: compare = (compare, None)
if type(ops) in types.StringTypes: ops = [ops]
base_dir = os.path.join(root_path, base[0])
compare_dir = os.path.join(root_path, compare[0])
if base[1] is None:
# base defaults to earliest capture
base = (base[0], max(os.listdir(base_dir)))
if compare[1] is None:
# compare defaults to latest capture
compare = (compare[0], min(os.listdir(compare_dir)))
out_path = os.path.join(out_path, base[0], base[1], compare[0], compare[1])
drivers.windowing.PreparePath(out_path)
# TODO(jhaas): right now we're just dumping output to a log file
# (and the console), which works as far as it goes but isn't nearly
# robust enough. Change this after deciding exactly what we want to
# change it to.
out_file = open(os.path.join(out_path, "log.txt"), "w")
description_string = ("Comparing %s %s to %s %s" %
(base[0], base[1], compare[0], compare[1]))
out_file.write(description_string)
print description_string
base_dir = os.path.join(base_dir, base[1])
compare_dir = os.path.join(compare_dir, compare[1])
for filename in os.listdir(base_dir):
out_file.write("%s: " % filename)
if not os.path.isfile(os.path.join(compare_dir, filename)):
out_file.write("Does not exist in target directory\n")
print "File %s does not exist in target directory" % filename
continue
base_filename = os.path.join(base_dir, filename)
compare_filename = os.path.join(compare_dir, filename)
for op in ops:
if type(op) in types.StringTypes: op = (op, None)
module = operators.GetOperator(op[0])
ret = module.Compare(base_filename, compare_filename)
if ret is None:
print "%s: OK" % (filename,)
out_file.write("OK\n")
else:
print "%s: %s" % (filename, ret[0])
out_file.write("%s\n" % (ret[0]))
ret[1].save(os.path.join(out_path, filename))
out_file.close()
def ExecuteCompare2(command):
"""Executes the Compare2 command."""
if command["--url"]:
url_list = [command["--url"]]
else:
startline = command["--startline"]
if command["--count"]:
endline = startline+command["--count"]
else:
endline = command["--endline"]
url_list = [url.strip() for url in
open(command["--list"], "r").readlines()[startline:endline]]
log_file = open(command["--logfile"], "w")
outdir = command["--outdir"]
if not outdir: outdir = tempfile.gettempdir()
scrape_info_list = []
class ScrapeInfo(object):
"""Helper class to hold information about a scrape."""
__slots__ = ["browser_path", "scraper", "outdir", "result"]
for index in xrange(1, 3):
scrape_info = ScrapeInfo()
scrape_info.browser_path = command["--browser%d" % index]
scrape_info.scraper = scrapers.GetScraper(
(command["--browser"], command["--browser%dver" % index]))
if command["--browser%dname" % index]:
scrape_info.outdir = os.path.join(outdir,
command["--browser%dname" % index])
else:
scrape_info.outdir = os.path.join(outdir, str(index))
drivers.windowing.PreparePath(scrape_info.outdir)
scrape_info_list.append(scrape_info)
compare = operators.GetOperator("equals_with_mask")
for url in url_list:
success = True
for scrape_info in scrape_info_list:
scrape_info.result = scrape_info.scraper.Scrape(
[url], scrape_info.outdir, command["--size"], (0, 0),
command["--timeout"], path=scrape_info.browser_path)
if not scrape_info.result:
scrape_info.result = "success"
else:
success = False
result = "unknown"
if success:
result = "equal"
file1 = drivers.windowing.URLtoFilename(
url, scrape_info_list[0].outdir, ".bmp")
file2 = drivers.windowing.URLtoFilename(
url, scrape_info_list[1].outdir, ".bmp")
comparison_result = compare.Compare(file1, file2,
maskdir=command["--maskdir"])
if comparison_result is not None:
result = "not-equal"
if command["--diffdir"]:
comparison_result[1].save(
drivers.windowing.URLtoFilename(url, command["--diffdir"], ".bmp"))
# TODO(jhaas): maybe use the logging module rather than raw file writes
log_file.write("%s %s %s %s\n" % (url,
scrape_info_list[0].result,
scrape_info_list[1].result,
result))