From fc27193898d644643167250730c51e07605f5d9c Mon Sep 17 00:00:00 2001
From: Saeed Rasooli <saeed.gnu@gmail.com>
Date: Fri, 3 Feb 2023 07:02:49 +0330
Subject: [PATCH] fix bug of glos.addEntryObj(dataEntry) adding empty file
 because tmpDataDir is not set until glos.read() set and create tmpDataDir on
 glos.tmpDataDir access, and add test, #424

---
 pyglossary/glossary.py       |  5 +++++
 pyglossary/glossary_utils.py |  6 +++---
 tests/glossary_test.py       | 24 ++++++++++++++++++++++++
 3 files changed, 32 insertions(+), 3 deletions(-)

diff --git a/pyglossary/glossary.py b/pyglossary/glossary.py
index c0788b4d5..b1d20f3ce 100644
--- a/pyglossary/glossary.py
+++ b/pyglossary/glossary.py
@@ -438,6 +438,8 @@ def filename(self) -> str:
 
 	@property
 	def tmpDataDir(self) -> str:
+		if not self._tmpDataDir:
+			self._setTmpDataDir(self._filename)
 		return self._tmpDataDir
 
 	def wordTitleStr(
@@ -532,6 +534,7 @@ def _createReader(self, format: str, options: "Dict[str, Any]") -> "Any":
 		return reader
 
 	def _setTmpDataDir(self, filename: str):
+		import uuid
 		# good thing about cacheDir is that we don't have to clean it up after
 		# conversion is finished.
 		# specially since dataEntry.save(...) will move the file from cacheDir
@@ -543,6 +546,8 @@ def _setTmpDataDir(self, filename: str):
 		# if self._hasWriteAccessToDir(f"{filename}_res", os.W_OK):
 		# 	self._tmpDataDir = f"{filename}_res"
 		# else:
+		if not filename:
+			filename = uuid.uuid1().hex
 		self._tmpDataDir = join(cacheDir, os.path.basename(filename) + "_res")
 		log.debug(f"tmpDataDir = {self._tmpDataDir}")
 		os.makedirs(self._tmpDataDir, mode=0o700, exist_ok=True)
diff --git a/pyglossary/glossary_utils.py b/pyglossary/glossary_utils.py
index ba340bba6..27b5f4082 100644
--- a/pyglossary/glossary_utils.py
+++ b/pyglossary/glossary_utils.py
@@ -40,10 +40,10 @@ def __init__(self, glos):
 		self._glos = glos
 		self._sortKey = None
 
-	def append(self, entry):
+	def append(self, entry: "BaseEntry"):
 		self._l.append(entry.getRaw(self._glos))
 
-	def insert(self, pos, entry):
+	def insert(self, pos, entry: "BaseEntry"):
 		self._l.insert(pos, entry.getRaw(self._glos))
 
 	def clear(self):
@@ -52,7 +52,7 @@ def clear(self):
 	def __len__(self):
 		return len(self._l)
 
-	def __iter__(self):
+	def __iter__(self) -> "Iterator[BaseEntry]":
 		glos = self._glos
 		for rawEntry in self._l:
 			yield Entry.fromRaw(
diff --git a/tests/glossary_test.py b/tests/glossary_test.py
index a4114be1e..9fb92d937 100644
--- a/tests/glossary_test.py
+++ b/tests/glossary_test.py
@@ -861,6 +861,30 @@ def test_addEntries_2(self):
 			[entry.l_word for entry in glos],
 		)
 
+	def test_addEntries_3(self):
+		glos = self.glos = Glossary()
+		glos.addEntryObj(glos.newEntry(["a"], "test 1"))
+		glos.addEntryObj(glos.newEntry(["b"], "test 3"))
+		glos.addEntryObj(glos.newDataEntry(
+			"file.bin",
+			b"hello\x00world",
+		))
+		glos.updateEntryFilters()
+		glos.updateIter()
+		wordListList = []
+		dataEntries = []
+		for entry in glos:
+			wordListList.append(entry.l_word)
+			if entry.isData():
+				dataEntries.append(entry)
+		self.assertEqual(
+			wordListList,
+			[['a'], ['b'], ["file.bin"]],
+		)
+		self.assertEqual(len(dataEntries), 1)
+		self.assertEqual(dataEntries[0].getFileName(), "file.bin")
+		self.assertEqual(dataEntries[0].data, b"hello\x00world")
+
 	def test_sortWords_1(self):
 		glos = self.glos = Glossary()
 		wordsList = self.addWords(