Identificação e inserção automática de xref no pipeline DOCX → SPS XML

markup_doc/labeling_utils.py

@@ -773,6 +773,17 @@ def create_labeled_object2(i, item, state, sections):
         obj["type"] = "paragraph"
         obj["value"] = {"label": state["label"], "paragraph": item.get("text")}
 
+    if state.get("body") and re.search(
+        r"^(refer[eê]nci|references?)\s*$", item.get("text").strip().lower()
+    ):
+        state["label"] = "<sec>"
+        state["body"] = False
+        state["back"] = True
+        result = {"label": "<sec>", "body": False, "back": True}
+        obj["type"] = "paragraph"
+        obj["value"] = {"label": state["label"], "paragraph": item.get("text")}
+
+
     if not result:
         result = {"label": "<p>", "body": state["body"], "back": state["back"]}
         state["label"] = result.get("label")
@@ -874,12 +885,12 @@ def get_data_first_block(text, metadata, user_id):
         "Content-Type": "application/json",
     }
 
+    resp_json = {}
     response = requests.post(url, json=payload, headers=headers)
 
     if response.status_code == 200:
         response_json = response.json()
         message_str = response_json["message"]
-
         resp_json = json.loads(message_str)
 
     return resp_json
@@ -1017,22 +1028,30 @@ def search_special_id(data_body, label):
             clean_label = re.sub(r"^[\s\.,;:–—-]+", "", label).capitalize()
 
             if d["type"] == "image":
-                if clean_label == data["figlabel"]:
-                    return data.get("figid")
+                figlabel = data.get("figlabel") or ""
+                figid = data.get("figid") or ""
+                if clean_label == figlabel:
+                    return figid or None
                 if (
-                    data["figid"][0] == clean_label.lower()[0]
-                    and data["figid"][1] in clean_label.lower()
+                    figid
+                    and len(figid) > 1
+                    and figid[0] == clean_label.lower()[:1]
+                    and figid[1] in clean_label.lower()
                 ):
-                    return data.get("figid")
+                    return figid
 
             if d["type"] == "table":
-                if clean_label == data["tablabel"]:
-                    return data.get("tabid")
+                tablabel = data.get("tablabel") or ""
+                tabid = data.get("tabid") or ""
+                if clean_label == tablabel:
+                    return tabid or None
                 if (
-                    data["tabid"][0] == clean_label.lower()[0]
-                    and data["tabid"][1] in clean_label.lower()
+                    tabid
+                    and len(tabid) > 1
+                    and tabid[0] == clean_label.lower()[:1]
+                    and tabid[1] in clean_label.lower()
                 ):
-                    return data.get("tabid")
+                    return tabid
 
     for d in data_body:
         if d["type"] in ["compound_paragraph"]:
@@ -1279,6 +1298,7 @@ def append_fragment(node_dest, val):
 
     clean = escape_angle_brackets_outside_tags(clean)
     clean = remove_unpaired_tags(clean)
+    clean = re.sub(r'<(?![/a-zA-Z_])', '&lt;', clean)
 
     if clean == "":
         parent = node_dest.getparent()
@@ -1351,23 +1371,17 @@ def proccess_special_content(text, data_body):
     res = []
     dict_type = {"f": "fig", "t": "table", "e": "disp-formula"}
 
-    try:
-        for match in re.finditer(
-            pattern, text, re.IGNORECASE | re.UNICODE | re.VERBOSE
-        ):
-            label = match.group(0)
-
-            id = search_special_id(data_body, label)
-
-            res.append(
-                {
-                    "label": label,
-                    "id": id,
-                    "reftype": dict_type.get(id[0].lower(), "other"),
-                }
-            )
-    except Exception as exc:
-        print(f"ERROR proccess_special_content: {exc}")
-        pass
+    for match in re.finditer(pattern, text, re.IGNORECASE | re.UNICODE | re.VERBOSE):
+        label = match.group(0)
+        id = search_special_id(data_body, label)
+        if id is None:
+            continue
+        res.append(
+            {
+                "label": label,
+                "id": id,
+                "reftype": dict_type.get(id[0].lower(), "other"),
+            }
+        )
 
     return res

markup_doc/migrations/0003_articledocxmarkup_marked_file.py

@@ -0,0 +1,18 @@
+# Generated by Django 6.0.5 on 2026-05-26 14:16
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('markup_doc', '0002_alter_articledocx_estatus_and_more'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='articledocxmarkup',
+            name='marked_file',
+            field=models.FileField(blank=True, null=True, upload_to='uploads_docx_marked/', verbose_name='Marked Document'),
+        ),
+    ]

markup_doc/migrations/0004_articledocxmarkup_xref_status.py

@@ -0,0 +1,16 @@
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('markup_doc', '0003_articledocxmarkup_marked_file'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='articledocxmarkup',
+            name='xref_status',
+            field=models.JSONField(blank=True, null=True, verbose_name='XRef Status'),
+        ),
+    ]

markup_doc/models.py

@@ -1,11 +1,14 @@
+import json
+import os
+
 from django import forms
 from django.db import models
 from django.db.models import Q
 from django.urls import reverse
-from django.utils.html import format_html
+from django.utils.html import format_html, mark_safe
 from django.utils.translation import gettext_lazy as _
 from modelcluster.models import ClusterableModel
-from wagtail.admin.panels import FieldPanel, ObjectList, TabbedInterface
+from wagtail.admin.panels import FieldPanel, ObjectList, Panel, TabbedInterface
 from wagtail.blocks import ChoiceBlock, StreamBlock, StructBlock, TextBlock
 from wagtail.fields import StreamField
 from wagtail.images.blocks import ImageChooserBlock
@@ -26,8 +29,6 @@ class ProcessStatus(models.IntegerChoices):
 class ReadOnlyFileWidget(forms.Widget):
     def render(self, name, value, attrs=None, renderer=None):
         if value:
-            # Muestra el archivo como un enlace de descarga
-            # return format_html('<a href="{}" target="_blank" download>{}</a>', value.url, value.name.split('/')[-1])
             instance = value.instance
             url = reverse("generate_xml", args=[instance.pk])
             return format_html(
@@ -36,6 +37,74 @@ def render(self, name, value, attrs=None, renderer=None):
         return ""
 
 
+class DownloadMarkedFileWidget(forms.Widget):
+    def render(self, name, value, attrs=None, renderer=None):
+        if value:
+            instance = value.instance
+            url = reverse("download_marked_docx", args=[instance.pk])
+            filename = os.path.basename(value.name)
+            return format_html(
+                '<a href="{}" target="_blank" download>{}</a>', url, filename
+            )
+        return mark_safe('<span style="color: gray;">Não disponível ainda</span>')
+
+
+class XrefStatusWidget(forms.Widget):
+    def render(self, name, value, attrs=None, renderer=None):
+        if isinstance(value, str):
+            try:
+                value = json.loads(value)
+            except (json.JSONDecodeError, ValueError):
+                value = None
+        if not value:
+            return mark_safe('<span style="color: gray;">Não processado</span>')
+        valid = value.get("valid", False)
+        total_refs = value.get("total_references", 0)
+        total_cits = value.get("total_citations", 0)
+        orphaned_bk = len(value.get("orphaned_bookmarks", []))
+        orphaned_hl = value.get("orphaned_hyperlinks", [])
+        color = "green" if valid else "red"
+        status = "Válido" if valid else "Inválido"
+        html = format_html(
+            '<p><strong style="color:{};">{}</strong> &nbsp; {} referências | {} citações linkadas</p>',
+            color, status, total_refs, total_cits,
+        )
+        if orphaned_hl:
+            html += format_html(
+                '<p style="color:red;">Citações sem referência: {}</p>',
+                ', '.join(orphaned_hl),
+            )
+        if orphaned_bk:
+            html += format_html(
+                '<p style="color:orange;">{} referência(s) sem citação no texto</p>',
+                orphaned_bk,
+            )
+        return html
+
+
+class ReprocessButtonPanel(Panel):
+    def __init__(self, confirm_message="Reprocessar este documento?", **kwargs):
+        super().__init__(**kwargs)
+        self.confirm_message = confirm_message
+
+    def clone_kwargs(self):
+        return {**super().clone_kwargs(), "confirm_message": self.confirm_message}
+
+    class BoundPanel(Panel.BoundPanel):
+        def render_html(self, parent_context=None):
+            if not self.instance or not self.instance.pk:
+                return ""
+            url = reverse("reprocess", args=[self.instance.pk])
+            msg = self.panel.confirm_message.replace("'", "\\'")
+            return mark_safe(
+                f'<div style="margin:12px 0;">'
+                f'<a href="{url}" onclick="return confirm(\'{msg}\')" '
+                f'style="padding:8px 16px;background:#e9a000;color:white;'
+                f'font-weight:bold;border-radius:4px;text-decoration:none;">'
+                f'Reprocessar</a></div>'
+            )
+
+
 class ArticleDocx(CommonControlField):
     title = models.TextField(_("Document Title"), null=True, blank=True)
     file = models.FileField(
@@ -347,6 +416,17 @@ class ArticleDocxMarkup(CommonControlField, ClusterableModel):
         verbose_name=_("Document"),
         upload_to="uploads_docx/",
     )
+    marked_file = models.FileField(
+        null=True,
+        blank=True,
+        verbose_name=_("Marked Document"),
+        upload_to="uploads_docx_marked/",
+    )
+    xref_status = models.JSONField(
+        _("XRef Status"),
+        null=True,
+        blank=True,
+    )
     estatus = models.IntegerField(
         _("Process estatus"),
         choices=ProcessStatus.choices,
@@ -448,6 +528,16 @@ def __str__(self):
         title = self.title or ""
         return f"{title} | {self.estatus}"
 
+    def get_marked_file_status(self):
+        if not self.marked_file:
+            return "Aguardando processamento"
+        if self.xref_status:
+            total = self.xref_status.get("total_references", 0)
+            cits = self.xref_status.get("total_citations", 0)
+            return f"✓ Disponível ({total} refs, {cits} citações)"
+        return "✓ Disponível"
+    get_marked_file_status.short_description = _("DOCX Marcado")
+
     @property
     def url_download(self):
         return self.file_xml.url if self.file_xml else None
@@ -510,6 +600,9 @@ class MarkupXML(ArticleDocxMarkup):
     panels_xml = [
         FieldPanel("file_xml", widget=ReadOnlyFileWidget()),
         FieldPanel("text_xml"),
+        ReprocessButtonPanel(
+            confirm_message="Isso irá descartar as edições manuais e reprocessar o DOCX original. Continuar?"
+        ),
     ]
 
     panels_details = [
@@ -551,3 +644,21 @@ class MarkupXML(ArticleDocxMarkup):
 
     class Meta:
         proxy = True
+
+
+class ProcessedDocx(ArticleDocxMarkup):
+    panels_doc = [
+        FieldPanel("title"),
+        FieldPanel("marked_file", widget=DownloadMarkedFileWidget()),
+        FieldPanel("xref_status", widget=XrefStatusWidget()),
+        ReprocessButtonPanel(confirm_message="Reprocessar este documento?"),
+    ]
+
+    edit_handler = TabbedInterface([
+        ObjectList(panels_doc, heading=_("DOCX Marcado")),
+    ])
+
+    class Meta:
+        proxy = True
+        verbose_name = _("DOCX processado")
+        verbose_name_plural = _("DOCXs processados")

markup_doc/static/js/xref-button.js

@@ -523,8 +523,76 @@ function get_zip() {
 
     // También en DOMContentLoaded por si basta
     document.addEventListener("DOMContentLoaded", tryAttach);
-  
+
     // Llama una vez por si ya está listo
     tryAttach();
   })();
 
+
+// Botão Reprocessar — aparece nas views de edição de ProcessedDocx e MarkupXML
+(function () {
+    var path = window.location.pathname;
+    var isProcessedDocx = path.indexOf('processeddocx/edit/') !== -1;
+    var isMarkupXml = path.indexOf('markupxml/edit/') !== -1;
+
+    if (!isProcessedDocx && !isMarkupXml) return;
+
+    var match = path.match(/\/edit\/(\d+)\//);
+    if (!match) return;
+    var pk = match[1];
+
+    function makeBtn() {
+        var btn = document.createElement('button');
+        btn.type = 'button';
+        btn.id = 'reprocess-btn';
+        btn.textContent = 'Reprocessar';
+        btn.style.cssText = [
+            'padding:4px 12px',
+            'cursor:pointer',
+            'background:#e9a000',
+            'color:white',
+            'font-weight:bold',
+            'border:none',
+            'border-radius:4px',
+            'margin-left:8px',
+            'font-size:14px',
+        ].join(';');
+        btn.addEventListener('mouseover', function () { btn.style.background = '#c98000'; });
+        btn.addEventListener('mouseout', function () { btn.style.background = '#e9a000'; });
+        btn.addEventListener('click', function () {
+            var msg = isMarkupXml
+                ? 'Isso irá descartar as edições manuais e reprocessar o DOCX original. Continuar?'
+                : 'Reprocessar este documento?';
+            if (confirm(msg)) {
+                window.location.href = '/admin/reprocess/' + pk + '/';
+            }
+        });
+        return btn;
+    }
+
+    function tryInsert() {
+        if (document.getElementById('reprocess-btn')) return true;
+        // Tenta área de ações do cabeçalho Wagtail (v5/v6)
+        var actionArea = document.querySelector('.w-slim-header__action-buttons')
+            || document.querySelector('[data-controller="w-slim-header"] .w-slim-header__title-wrapper');
+        if (actionArea) {
+            actionArea.appendChild(makeBtn());
+            return true;
+        }
+        // Fallback: insere após o primeiro botão submit (salvar)
+        var saveBtn = document.querySelector('button[type="submit"]');
+        if (saveBtn && saveBtn.parentNode) {
+            saveBtn.parentNode.insertBefore(makeBtn(), saveBtn.nextSibling);
+            return true;
+        }
+        return false;
+    }
+
+    if (!tryInsert()) {
+        var obs = new MutationObserver(function () {
+            if (tryInsert()) obs.disconnect();
+        });
+        obs.observe(document.documentElement, { childList: true, subtree: true });
+    }
+})();
+