Resolve long tree string problem on Windows #258

etetoolkit · May 23, 2019 · 1752154 · 1752154
1 parent 6f3d59c
commit 1752154
Show file tree

Hide file tree

Showing 2 changed files with 59 additions and 2 deletions.
diff --git a/ete3/parser/newick.py b/ete3/parser/newick.py
@@ -230,8 +230,20 @@ class as the root(This allows to work with custom TreeNode
         from ..coretype.tree import TreeNode
         root_node = TreeNode()
 
-    if isinstance(newick, six.string_types):   
-        if os.path.exists(newick):
+    if isinstance(newick, six.string_types):
+
+        # try to determine whether the file exists.  
+        # For very large trees, if newick contains the content of the tree, rather than a file name,
+        # this may fail, at least on Windows, because the os fails to stat the file content, deeming it 
+        # too long for testing with os.path.exists.  This raises a ValueError with description
+        # "stat: path too long for Windows".  This is described in issue #258
+        try: 
+            file_exists = os.path.exists(newick)
+        except ValueError:      # failed to stat
+            file_exists = False
+
+        # if newick refers to a file, read it from file; otherwise, regard it as a Newick content string.
+        if file_exists:
             if newick.endswith('.gz'):
                 import gzip
                 nw = gzip.open(newick).read()
@@ -240,6 +252,7 @@ class as the root(This allows to work with custom TreeNode
         else:
             nw = newick
 
+
         matcher = compile_matchers(formatcode=format)
         nw = nw.strip()
         if not nw.startswith('(') and nw.endswith(';'):

diff --git a/ete3/test/test_issue258/issue258.py b/ete3/test/test_issue258/issue258.py