Merge pull request #1384 from mathics/fix1379-rebase

Make Python package lxml optional

Author: rocky

mathics/builtin/base.py

@@ -23,8 +23,8 @@
     PrecisionReal,
     String,
     Symbol,
-    SymbolTrue,
     SymbolFalse,
+    SymbolTrue,
     ensure_context,
     strip_context,
 )

mathics/builtin/fileformats/htmlformat.py

@@ -17,13 +17,13 @@
 from mathics.builtin.base import MessageException
 
 from io import BytesIO
+import platform
 import re
-import sys
+
 
 try:
     import lxml.html as lhtml
 except ImportError:
-    print("lxml.html is not available...")
     pass
 
 
@@ -86,13 +86,12 @@ class ParseError(Exception):
     pass
 
 
-if "__pypy__" in sys.builtin_module_names:
+if platform.python_implementation() == "PyPy":
 
     def parse_html_stream(f):
         parser = lhtml.HTMLParser(encoding="utf8")
         return lhtml.parse(f, parser)
 
-
 else:
 
     def parse_html_stream(f):
@@ -249,7 +248,7 @@ def traverse(parent):
 class DataImport(_DataImport):
     """
     >> Import["ExampleData/PrimeMeridian.html", "Data"][[1, 1, 2, 3]]
-     = {Washington, D.C., 77°03′56.07″ W (1897) or 77°04′02.24″ W (NAD 27) or 77°04′01.16″ W (NAD 83), New Naval Observatory meridian}
+     = {Washington, D.C., 77...03′56.07″ W (1897) or 77...04′02.24″ W (NAD 27) or 77...04′01.16″ W (NAD 83), New Naval Observatory meridian}
 
     #> Length[Import["ExampleData/PrimeMeridian.html", "Data"]]
      = 3

mathics/builtin/inout.py

@@ -1930,7 +1930,7 @@ class General(Builtin):
         "syntax": "`1`",
         "invalidargs": "Invalid arguments.",
         "notboxes": "`1` is not a valid box structure.",
-        "pyimport": '`1`[] is not available. Your Python installation misses the "`2`" module.',
+        "pyimport": '`1`[] is not available. Python module "`2`" is not installed.',
     }
 
 

mathics/builtin/lists.py

@@ -1117,6 +1117,8 @@ def apply_makeboxes(self, list, i, f, evaluation):
     def apply(self, list, i, evaluation):
         "Part[list_, i___]"
 
+        if list == SymbolFailed:
+            return
         indices = i.get_sequence()
         # How to deal with ByteArrays
         if list.get_head_name() == "System`ByteArray":

setup.py

@@ -102,7 +102,9 @@ def read(*rnames):
     "requests",
     "scikit-image",
     "wordcloud",  # Used in builtin/image.py by WordCloud()
-    "lxml", # Used in builtin/fileformats/html
+
+    # lxml is an optional dependency for HTML parsing
+    # "lxml", # Used in builtin/fileformats/html
 ]