libextractor

gen_webp_testdata.sh (5367B)

---

 #!/bin/sh
 # This file is part of libextractor.
 # Copyright (C) 2026 Vidyut Samanta and Christian Grothoff
 #
 # Generate the test data for the "webp" plugin:
 #
 #   src/plugins/testdata/webp_test.webp
 #
 # An extended (VP8X) animated WebP with an alpha channel, an ICC
 # profile, three animation frames and EXIF and XMP chunks -- that is,
 # every optional feature the container defines, so that one file
 # exercises the whole chunk walk.
 #
 # The three frames carry a genuine lossless VP8L bitstream, encoded here
 # by ffmpeg from a solid colour generated by lavfi (so nothing is read
 # from disk and the result does not depend on a sample image).  The
 # container around it is assembled by hand because no encoder emits all
 # of these chunks at once.  The ICC, EXIF and XMP payloads are
 # placeholders: the plugin only reports that they are present, and
 # parsing them is the exiv2 plugin's job.
 #
 # Needs: ffmpeg built with libwebp (Debian package "ffmpeg") and
 # python3.  Deterministic: no timestamps, no random data.
 #
 # Written to CC0; there is no third party content in this file.
 #
 # Usage: contrib/gen_webp_testdata.sh [output-directory]
 
 set -e
 
 OUT="${1:-src/plugins/testdata}"
 mkdir -p "$OUT"
 
 TMP=$(mktemp -d)
 trap 'rm -rf "$TMP"' EXIT
 
 # A 40x24 solid colour, encoded losslessly.  libwebp emits a bare
 # "RIFF....WEBPVP8L" file, whose VP8L chunk we lift out below.
 ffmpeg -hide_banner -loglevel error -y \
        -f lavfi -i "color=c=0x1EB45A:s=40x24" \
        -frames:v 1 -c:v libwebp -lossless 1 -pix_fmt bgra \
        "$TMP/frame.webp"
 
 python3 - "$OUT" "$TMP/frame.webp" <<'EOF'
 import struct
 import sys
 
 out = sys.argv[1]
 frame_path = sys.argv[2]
 
 WIDTH = 40
 HEIGHT = 24
 
 # VP8X feature flags, in the single flag byte at the start of the
 # "VP8X" payload: Rsv(2) ICC(1) Alpha(1) EXIF(1) XMP(1) Anim(1) Rsv(1)
 ICC_FLAG = 0x20
 ALPHA_FLAG = 0x10
 EXIF_FLAG = 0x08
 XMP_FLAG = 0x04
 ANIM_FLAG = 0x02
 
 
 def chunk(fourcc, payload):
     """A RIFF chunk: 4CC, 32-bit little-endian size, payload, pad byte."""
     assert len(fourcc) == 4
     data = fourcc.encode('ascii') + struct.pack('<I', len(payload)) + payload
     if len(payload) & 1:
         data += b'\0'
     return data
 
 
 def le24(value):
     return struct.pack('<I', value)[:3]
 
 
 def extract_vp8l(path):
     """Lift the VP8L chunk (header included) out of a simple WebP file."""
     with open(path, 'rb') as f:
         data = f.read()
     assert data[0:4] == b'RIFF' and data[8:12] == b'WEBP', 'not a WebP file'
     pos = 12
     while pos + 8 <= len(data):
         fourcc = data[pos:pos + 4]
         size = struct.unpack('<I', data[pos + 4:pos + 8])[0]
         if fourcc == b'VP8L':
             end = pos + 8 + size + (size & 1)
             return data[pos:end]
         pos += 8 + size + (size & 1)
     raise SystemExit('ffmpeg did not produce a VP8L chunk')
 
 
 vp8l = extract_vp8l(frame_path)
 
 # A placeholder ICC profile: a 128 byte header saying "RGB display
 # profile, no tags".  Small, self-describing, and never parsed by us.
 icc = (struct.pack('>I', 132)
        + b'lcms' + struct.pack('>I', 0x04300000)
        + b'mntr' + b'RGB ' + b'XYZ '
        + b'\0' * 12                      # creation date, zeroed
        + b'acsp' + b'APPL' + b'\0' * 4
        + b'\0' * 12                      # device manufacturer/model
        + b'\0' * 8                       # device attributes
        + b'\0' * 4                       # rendering intent: perceptual
        + struct.pack('>III', 0xF6D6, 0x10000, 0xD32D)   # D50 white point
        + b'none' + b'\0' * 44
        + struct.pack('>I', 0))           # tag count: zero
 
 # "Exif\0\0" is not part of the WebP EXIF chunk; the payload is the raw
 # TIFF structure.  A little-endian TIFF header with an empty IFD.
 exif = b'II' + struct.pack('<HI', 42, 8) + struct.pack('<H', 0) \
     + struct.pack('<I', 0)
 
 xmp = (b'<?xpacket begin="" id="W5M0MpCehiHzreSzNTczkc9d"?>'
        b'<x:xmpmeta xmlns:x="adobe:ns:meta/">'
        b'<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">'
        b'<rdf:Description rdf:about="" xmlns:dc="http://purl.org/dc/elements/1.1/">'
        b'<dc:title><rdf:Alt><rdf:li xml:lang="x-default">'
        b'libextractor webp test</rdf:li></rdf:Alt></dc:title>'
        b'</rdf:Description></rdf:RDF></x:xmpmeta>'
        b'<?xpacket end="w"?>')
 
 vp8x = chunk('VP8X',
              bytes([ICC_FLAG | ALPHA_FLAG | EXIF_FLAG | XMP_FLAG | ANIM_FLAG])
              + b'\0\0\0'
              + le24(WIDTH - 1)
              + le24(HEIGHT - 1))
 
 # Background colour is BGRA; a loop count of zero means "forever".
 anim = chunk('ANIM', struct.pack('<IH', 0xFF204080, 0))
 
 # Frame durations of 100, 150 and 250 ms sum to a round 500 ms.
 frames = b''
 for duration in (100, 150, 250):
     frames += chunk('ANMF',
                     le24(0) + le24(0)                    # frame_x, frame_y
                     + le24(WIDTH - 1) + le24(HEIGHT - 1)
                     + le24(duration)
                     + bytes([0x02])                      # dispose, no blend
                     + vp8l)
 
 body = (b'WEBP'
         + vp8x
         + chunk('ICCP', icc)
         + anim
         + frames
         + chunk('EXIF', exif)
         + chunk('XMP ', xmp))
 
 data = b'RIFF' + struct.pack('<I', len(body)) + body
 
 path = out + '/webp_test.webp'
 with open(path, 'wb') as f:
     f.write(data)
 print('%s: %d bytes, ICC %d bytes' % (path, len(data), len(icc)))
 EOF