Faircot process

modules/automation/automation.py

@@ -10,6 +10,7 @@ import io
 import logging
 import json
 import re
+import traceback
 
 logger = logging.getLogger(__name__)
 
@@ -62,25 +63,57 @@ class AutomationDocument(ModelSQL, ModelView, Workflow):
     @ModelView.button
     def run_ocr(cls, docs):
         for doc in docs:
+            logger.info(
+                "RUN_OCR_START doc=%s type=%s state=%s incoming=%s",
+                getattr(doc, 'id', None), getattr(doc, 'type', None),
+                getattr(doc, 'state', None),
+                getattr(getattr(doc, 'document', None), 'id', None))
             try:
                 if doc.type == 'weight_report':
                     # Décoder le fichier depuis le champ Binary
-                    file_data = doc.document.data or b""
+                    if not doc.document:
-                    logger.info(f"File size: {len(file_data)} bytes")
+                        raise ValueError("No incoming document linked")
-                    logger.info(f"First 20 bytes: {file_data[:20]}")
-                    logger.info(f"Last 20 bytes: {file_data[-20:]}")
 
+                    file_data = doc.document.data or b""
                     file_name = doc.document.name or "document"
+                    url = "http://automation-service:8006/ocr"
+                    logger.info(
+                        "RUN_OCR_FILE doc=%s incoming=%s name=%s size=%s "
+                        "first_20=%r last_20=%r",
+                        getattr(doc, 'id', None), doc.document.id,
+                        file_name, len(file_data), file_data[:20],
+                        file_data[-20:])
+
+                    if not file_data:
+                        raise ValueError("Incoming document has no binary data")
 
                     # Envoyer le fichier au service OCR
+                    logger.info(
+                        "RUN_OCR_HTTP_REQUEST doc=%s url=%s file_name=%s "
+                        "size=%s",
+                        getattr(doc, 'id', None), url, file_name,
+                        len(file_data))
                     response = requests.post(
-                        "http://automation-service:8006/ocr",
+                        url,
-                        files={"file": (file_name, io.BytesIO(file_data))}
+                        files={"file": (file_name, io.BytesIO(file_data))},
+                        timeout=120,
                     )
+                    logger.info(
+                        "RUN_OCR_HTTP_RESPONSE doc=%s status=%s reason=%s "
+                        "elapsed=%s content_type=%s body_start=%r",
+                        getattr(doc, 'id', None), response.status_code,
+                        response.reason, response.elapsed,
+                        response.headers.get('content-type'),
+                        response.text[:1000])
                     response.raise_for_status()
                     data = response.json()
-                    logger.info("RUN_OCR_RESPONSE:%s",data)
+                    ocr_text = data.get("ocr_text", "") or ""
-                    doc.ocr_text = data.get("ocr_text", "")
+                    logger.info(
+                        "RUN_OCR_JSON doc=%s keys=%s ocr_text_len=%s "
+                        "ocr_text_start=%r",
+                        getattr(doc, 'id', None), list(data.keys()),
+                        len(ocr_text), ocr_text[:500])
+                    doc.ocr_text = ocr_text
                     doc.state = "ocr_done"
                     doc.notes = (doc.notes or "") + "OCR done\n"
                 else:
@@ -112,9 +145,17 @@ class AutomationDocument(ModelSQL, ModelView, Workflow):
                                 doc.notes = (doc.notes or "") + " SO updated"
 
             except Exception as e:
+                logger.error(
+                    "RUN_OCR_FAILED doc=%s type=%s error=%s traceback=%s",
+                    getattr(doc, 'id', None), getattr(doc, 'type', None),
+                    e, traceback.format_exc())
                 doc.state = "error"
                 doc.notes = (doc.notes or "") + f"OCR error: {e}\n"
             doc.save()
+            logger.info(
+                "RUN_OCR_END doc=%s state=%s ocr_text_len=%s",
+                getattr(doc, 'id', None), getattr(doc, 'state', None),
+                len(getattr(doc, 'ocr_text', None) or ""))
     # -------------------------------------------------------
     #                   STRUCTURE (doctr)
     # -------------------------------------------------------

modules/document_incoming_wr/document.py

@@ -7,6 +7,9 @@ from trytond.modules.document_incoming.exceptions import (
     DocumentIncomingProcessError)
 from trytond.pool import Pool, PoolMeta
 import json
+import logging
+
+logger = logging.getLogger(__name__)
 
 
 class IncomingConfiguration(metaclass=PoolMeta):
@@ -39,12 +42,22 @@ class Incoming(metaclass=PoolMeta):
         return super()._get_results() | {'automation.document'}
 
     def _process_weight_report(self):
+        logger.info(
+            "DOCUMENT_INCOMING_WR_START incoming=%s name=%s type=%s "
+            "data_size=%s",
+            getattr(self, 'id', None), getattr(self, 'name', None),
+            getattr(self, 'type', None),
+            len(getattr(self, 'data', None) or b""))
         WR = Pool().get('automation.document')
         wr = WR()
         wr.document = self.id
         wr.type = 'weight_report'
         wr.state = 'draft'
         WR.save([wr])
+        logger.info(
+            "DOCUMENT_INCOMING_WR_AUTOMATION_CREATED incoming=%s "
+            "automation_document=%s",
+            getattr(self, 'id', None), getattr(wr, 'id', None))
         WR.run_ocr([wr])
         WR.run_metadata([wr])
         WR.run_pipeline([wr])