added comment

This commit is contained in:
2026-07-17 17:56:23 +02:00
parent 0d56cfe523
commit 15e827fbaa

View File

@ -1,4 +1,41 @@
#!/bin/bash #!/bin/bash
# -----------------------------------------------------------------------------
# WCX Thumbnail OCR
#
# Extracts text from a local image file or an image URL using the Google Cloud
# Vision API.
#
# The script:
# 1. Accepts either a local image path or an HTTP/HTTPS image URL.
# 2. Reads or downloads the image and encodes it as Base64.
# 3. Sends the image to Google Cloud Vision using TEXT_DETECTION.
# 4. Extracts the first detected text annotation from the API response.
# 5. Writes the OCR text to standard output or to a file specified with -o.
#
# Requirements:
# - GOOGLE_VISION_API_KEY must be set in the environment.
# - curl is required for HTTP requests and remote image downloads.
# - base64 is required for encoding the image.
# - jq is recommended and is required for clean plain-text output.
#
# Usage:
# ocr.sh <image-file-or-url> [-o output-file]
#
# Examples:
# ./ocr.sh thumbnail.jpg
# ./ocr.sh "https://example.com/thumbnail.jpg"
# ./ocr.sh "https://example.com/thumbnail.jpg" -o ocr-result.txt
#
# Output contract:
# - Without -o, the detected OCR text is written to standard output.
# - With -o, the OCR text is written to the specified file and a completion
# message is written to standard output.
# - A non-zero exit status indicates invalid input, missing configuration,
# missing dependencies, an API failure, or that no text was detected.
#
# The output is raw OCR text. Its structure is not guaranteed by Google Vision
# and must be parsed and validated separately before being stored as metadata.
# -----------------------------------------------------------------------------
# --- KONFIGURATION --- # --- KONFIGURATION ---
# Skriptet förväntar sig nu miljövariabeln: GOOGLE_VISION_API_KEY # Skriptet förväntar sig nu miljövariabeln: GOOGLE_VISION_API_KEY