Adicionar app/io/document.py

app/io/document.py

@@ -0,0 +1,72 @@
+# All comments in English.
+
+from docx import Document
+from odf.opendocument import load
+from odf.text import P
+from bs4 import BeautifulSoup
+import markdown
+
+def normalize_newlines(value):
+    if value is None:
+        return ""
+    return (
+        str(value)
+        .replace("\r\n", "\\n")
+        .replace("\n\r", "\\n")
+        .replace("\r", "\\n")
+        .replace("\n", "\\n")
+    )
+
+def convert_docx(path):
+    doc = Document(path)
+    lines = []
+    for p in doc.paragraphs:
+        lines.append(normalize_newlines(p.text))
+    return "\n".join(lines)
+
+def convert_doc(path):
+    # DOC (binary) requires LibreOffice headless conversion
+    import subprocess, tempfile, os
+    tmp = tempfile.mktemp(suffix=".docx")
+    subprocess.run(["soffice", "--headless", "--convert-to", "docx", "--outdir", "/tmp", path])
+    return convert_docx(tmp)
+
+def convert_odt(path):
+    doc = load(path)
+    lines = []
+    for p in doc.getElementsByType(P):
+        text = p.firstChild.data if p.firstChild else ""
+        lines.append(normalize_newlines(text))
+    return "\n".join(lines)
+
+def convert_txt(path):
+    with open(path, "r", encoding="utf-8", errors="ignore") as f:
+        return normalize_newlines(f.read())
+
+def convert_md(path):
+    with open(path, "r", encoding="utf-8", errors="ignore") as f:
+        raw = f.read()
+    html = markdown.markdown(raw)
+    soup = BeautifulSoup(html, "html.parser")
+    return normalize_newlines(soup.get_text())
+
+def convert_html(path):
+    with open(path, "r", encoding="utf-8", errors="ignore") as f:
+        raw = f.read()
+    soup = BeautifulSoup(raw, "html.parser")
+    return normalize_newlines(soup.get_text())
+
+def convert_document(path):
+    if path.endswith(".docx"):
+        return convert_docx(path)
+    if path.endswith(".doc"):
+        return convert_doc(path)
+    if path.endswith(".odt"):
+        return convert_odt(path)
+    if path.endswith(".txt"):
+        return convert_txt(path)
+    if path.endswith(".md"):
+        return convert_md(path)
+    if path.endswith(".html") or path.endswith(".htm"):
+        return convert_html(path)
+    raise ValueError("Unsupported document format")