Merge pull request #552 from girder/test-utility-scripts

Add some utility scripts that have been useful in testing.

test/utils/README.rst

@@ -0,0 +1,16 @@
+********************************
+Large Image Taet Utility Scripts
+********************************
+
+This directory consists of files that were useful in testing various aspects of
+large_image.  These files are not part of the stable library interface nor part
+of the regularly used testing code.  They may not be maintained.
+
+Some utilities:
+
+- compression_test.py - Recompress input files with a wide set of options.
+  Compute full statistics on each output file to determine the amount of loss
+  introduced with those compression options when appropriate.
+
+- compression_test_summary.py - Collect the results from the compression test
+  and output a csv file.

test/utils/compression_test.py

@@ -0,0 +1,80 @@
+import concurrent.futures
+import itertools
+import os
+import psutil
+import subprocess
+import sys
+
+os.environ['GDAL_PAM_PROXY_DIR'] = '/tmp/gdal'
+
+allOpts = ['-w', '--stats-full']
+matrix = [[
+    [],
+    ['--compression', 'none'],
+    ['--compression', 'jpeg'],
+    ['--compression', 'jpeg', '-q', '95'],
+    ['--compression', 'jpeg', '-q', '90'],
+    ['--compression', 'jpeg', '-q', '80'],
+    ['--compression', 'jpeg', '-q', '70'],
+    ['--compression', 'deflate'],
+    ['--compression', 'deflate', '--level', '1'],
+    ['--compression', 'deflate', '--level', '9'],
+    ['--compression', 'lzw'],
+    ['--compression', 'lzw', '--predictor', 'none'],
+    ['--compression', 'lzw', '--predictor', 'horizontal'],
+    ['--compression', 'zstd'],
+    ['--compression', 'zstd', '--level', '1'],
+    ['--compression', 'zstd', '--level', '9'],
+    ['--compression', 'zstd', '--level', '22'],
+    ['--compression', 'packbits'],
+    # ['--compression', 'jbig'],
+    # ['--compression', 'lzma'],
+    ['--compression', 'webp'],
+    ['--compression', 'webp', '-q', '0'],
+    ['--compression', 'webp', '-q', '100'],
+    ['--compression', 'webp', '-q', '95'],
+    ['--compression', 'webp', '-q', '90'],
+    ['--compression', 'webp', '-q', '80'],
+    ['--compression', 'webp', '-q', '70'],
+    ['--compression', 'jp2k'],
+    ['--compression', 'jp2k', '--psnr', '80'],
+    ['--compression', 'jp2k', '--psnr', '70'],
+    ['--compression', 'jp2k', '--psnr', '60'],
+    ['--compression', 'jp2k', '--psnr', '50'],
+    ['--compression', 'jp2k', '--psnr', '40'],
+    ['--compression', 'jp2k', '--cr', '10'],
+    ['--compression', 'jp2k', '--cr', '100'],
+    ['--compression', 'jp2k', '--cr', '1000'],
+], [
+    [],  # 256
+    ['--tile', '512'],
+    ['--tile', '1024'],
+]]
+
+if not len(sys.argv[1:]) or '--help' in sys.argv[1:]:
+    print("""test_compression.py [(concurrency)] (output directory) (input file ...)""")
+    sys.exit(0)
+args = sys.argv[1:]
+concurrency = psutil.cpu_count(logical=True)
+if args[0].isdigit():
+    concurrency = int(args[0])
+    args = args[1:]
+# Set to 1 to disable concurrency
+pool = concurrent.futures.ThreadPoolExecutor(max_workers=concurrency)
+tasks = []
+for input in args[1:]:
+    root = os.path.join(args[0], os.path.basename(input))
+    for optList in itertools.product(*matrix):
+        opts = [opt for subList in optList for opt in subList]
+        output = root + '.' + '.'.join(str(opt).strip('-') for opt in opts) + '.tiff'
+        output = output.replace('..', '.')
+        cmd = ['large_image_converter', input, output] + opts + allOpts
+        tasks.append((cmd, pool.submit(subprocess.call, cmd)))
+while len(tasks):
+    try:
+        tasks[0][-1].result(0.1)
+    except concurrent.futures.TimeoutError:
+        continue
+    cmd, task = tasks.pop(0)
+    print(' '.join(cmd))
+pool.shutdown(False)

test/utils/compression_test_summary.py

@@ -0,0 +1,48 @@
+import json
+import os
+import pandas
+import sys
+import tifftools
+
+if not len(sys.argv[1:]) or '--help' in sys.argv[1:]:
+    print("""test_compression_summary.py (directory) (csv output file)
+
+Check each file in the specified directory and gather appropriate statistics
+into csv output.""")
+    sys.exit(0)
+results = []
+for file in os.listdir(sys.argv[1]):
+    path = os.path.join(sys.argv[1], file)
+    entry = {'file': file}
+    if '.compression' in file:
+        entry['source'] = file.split('.compression')[0]
+    elif '.tile' in file:
+        entry['source'] = file.split('.tile')[0]
+    else:
+        entry['source'] = file.split('.tiff')[0]
+    entry['error'] = None
+    results.append(entry)
+    try:
+        info = tifftools.read_tiff(path)
+    except Exception:
+        entry['error'] = "Can't read"
+        print(entry)
+        continue
+    try:
+        details = json.loads(info['ifds'][0]['tags'][tifftools.Tag.ImageDescription.value]['data'])
+    except Exception:
+        entry['error'] = "Can't parse ImageDescription"
+        print(entry)
+        continue
+    entry.update(details['large_image_converter']['arguments'])
+    try:
+        entry.update({
+            k if k != 'psnr' else 'stats_psnr': v
+            for k, v in details['large_image_converter']['conversion_stats'].items()})
+    except Exception:
+        entry['error'] = 'No conversion stats'
+        print(entry)
+        continue
+    print(entry)
+df = pandas.DataFrame.from_dict(results)
+df.to_csv(sys.argv[2], index=False)