Fixes the bug identified in https://progresssoftware.atlassian.net/browse/MLE-14411

Python Client throws error when "filename" is not returned in response from the REST Api.
improvement

Author: BillFarber

CONTRIBUTING.md

@@ -76,6 +76,18 @@ project's documentation:
 
     python -i shell/docs.py
 
+
+## Testing updates in a different local project
+If you are using this in another project and making changes for it, use the following command to make the
+changes to this local project immediately reflected in a dependent project:
+```poetry add <local-path-to-this-project>/marklogic-python-client/```
+ 
+Using this method will allow you to very easily test changes to this project, in a different local project.
+
+Keep in mind that you probably do not want to check that version of the pyproject.toml file into version
+control since it is only useful locally.
+
+
 ## Testing the documentation locally
 
 The docs for this project are stored in the `./docs` directory as a set of Markdown files. These are published via

marklogic/documents.py

@@ -1,5 +1,6 @@
 import json
 from collections import OrderedDict
+from email.message import Message
 from typing import Union
 
 from marklogic.transactions import Transaction
@@ -262,23 +263,29 @@ def _extract_values_from_header(part) -> dict:
     Returns a dict containing values about the document content or metadata.
     """
     encoding = part.encoding
-    disposition = part.headers["Content-Disposition".encode(encoding)].decode(encoding)
-    disposition_values = {}
-    for item in disposition.split(";"):
-        tokens = item.split("=")
-        # The first item will be "attachment" and can be ignored.
-        if len(tokens) == 2:
-            disposition_values[tokens[0].strip()] = tokens[1]
+    disposition = part.headers["Content-Disposition".encode(encoding)].decode(
+        encoding
+    )
 
     content_type = None
     if part.headers.get("Content-Type".encode(encoding)):
-        content_type = part.headers["Content-Type".encode(encoding)].decode(encoding)
+        content_type = part.headers["Content-Type".encode(encoding)].decode(
+            encoding
+        )
 
-    uri = disposition_values["filename"]
-    if uri.startswith('"'):
-        uri = uri[1:]
-    if uri.endswith('"'):
-        uri = uri[:-1]
+    content_disposition_header = part.headers[
+        "Content-Disposition".encode(encoding)
+    ].decode(encoding)
+    msg = Message()
+    msg["content-disposition"] = content_disposition_header
+    uri = msg.get_filename()
+
+    disposition_values = {}
+    for item in disposition.replace(uri, "").split(";"):
+        tokens = item.split("=")
+        key = tokens[0].strip()
+        if key in ["category", "versionId"]:
+            disposition_values[key] = tokens[1]
 
     return {
         "uri": uri,

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "marklogic-python-client"
-version = "1.1.0"
+version = "1.1.1"
 description = "Python client for MarkLogic, built on the requests library"
 authors = ["MarkLogic <[email protected]>"]
 readme = "README.md"

test-app/.gitignore

@@ -1,3 +1,4 @@
 .gradle
 gradle-local.properties
 build
+docker

test-app/docker-compose.yml

@@ -0,0 +1,18 @@
+version: '3.8'
+name: marklogic_python
+
+services:
+
+  marklogic:
+    image: "marklogicdb/marklogic-db:11.2.0-centos-1.1.2"
+    platform: linux/amd64
+    environment:
+      - INSTALL_CONVERTERS=true
+      - MARKLOGIC_INIT=true
+      - MARKLOGIC_ADMIN_USERNAME=admin
+      - MARKLOGIC_ADMIN_PASSWORD=admin
+    volumes:
+      - ./docker/marklogic/logs:/var/opt/MarkLogic/Logs
+    ports:
+      - "8000-8002:8000-8002"
+      - "8030-8031:8030-8031"

test-app/src/main/ml-data/doc2;copy.xml

@@ -0,0 +1 @@
+<hello>semicolon</hello>

test-app/src/main/ml-data/doc2=copy.xml

@@ -0,0 +1 @@
+<hello>equal</hello>

tests/test_search_docs.py

@@ -84,10 +84,11 @@ def test_search_with_original_response(client: Client):
 
 
 def test_collection(client: Client):
+
     docs = client.documents.search(
         categories=["content", "collections"], collections=["search-test"]
     )
-    assert len(docs) == 2
+    assert len(docs) == 4
 
     doc1 = next(doc for doc in docs if doc.uri == "/doc1.json")
     assert doc1.content is not None
@@ -101,6 +102,18 @@ def test_collection(client: Client):
     assert "test-data" in doc1.collections
     assert "search-test" in doc1.collections
 
+    doc3 = next(doc for doc in docs if doc.uri == "/doc2;copy.xml")
+    assert doc3.content is not None
+    assert len(doc3.collections) == 2
+    assert "test-data" in doc3.collections
+    assert "search-test" in doc3.collections
+
+    doc4 = next(doc for doc in docs if doc.uri == "/doc2=copy.xml")
+    assert doc4.content is not None
+    assert len(doc4.collections) == 2
+    assert "test-data" in doc4.collections
+    assert "search-test" in doc4.collections
+
 
 def test_not_rest_user(not_rest_user_client: Client):
     response: Response = not_rest_user_client.documents.search(q="hello")
@@ -109,3 +122,27 @@ def test_not_rest_user(not_rest_user_client: Client):
     ), """The user does not have the rest-reader privilege, so MarkLogic is expected
     to return a 403. And the documents.search method is then expected to return the
     Response so that the user has access to everything in it."""
+
+
+def test_version_id(client: Client):
+    equalSignEtag = (
+        client.get("v1/documents?uri=/doc2=copy.xml")
+        .headers["ETag"]
+        .replace('"', "")
+    )
+
+    semicolonEtag = (
+        client.get("v1/documents?uri=/doc2;copy.xml")
+        .headers["ETag"]
+        .replace('"', "")
+    )
+
+    docs = client.documents.search(
+        categories=["content", "collections"], collections=["search-test"]
+    )
+
+    doc1 = next(doc for doc in docs if doc.uri == "/doc2=copy.xml")
+    assert doc1.version_id == equalSignEtag
+
+    doc2 = next(doc for doc in docs if doc.uri == "/doc2;copy.xml")
+    assert doc2.version_id == semicolonEtag