#8 Work in progress:

-Check XML schemas during parsing
FABallemand · Aug 21, 2023 · c51888d · c51888d
1 parent e9d5371
commit c51888d
Show file tree

Hide file tree

Showing 6 changed files with 1,073 additions and 11 deletions.
diff --git a/ezgpx/gpx_parser/parser.py b/ezgpx/gpx_parser/parser.py
@@ -1,7 +1,9 @@
+import os
 from typing import Optional, Union
 import logging
 from datetime import datetime
 import xml.etree.ElementTree as ET
+import xmlschema
 
 from ..gpx_elements import Bounds, Copyright, Email, Extensions, Gpx, Link, Metadata, Person, Point, PointSegment, Route, TrackSegment, Track, WayPoint
 
@@ -13,14 +15,16 @@ class Parser():
     GPX file parser.
     """
 
-    def __init__(self, file_path: str = "") -> None:
+    def __init__(self, file_path: Optional[str] = None) -> None:
         """
         initialize Parser instance.
 
         Args:
-            file_path (str, optional): Path to the file to parse. Defaults to "".
+            file_path (str, optional): Path to the file to parse. Defaults to None.
         """
         self.file_path: str = file_path
+        if not os.path.exists(self.file_path):
+            logging.warning("File path does not exist")
 
         self.gpx_tree: ET.ElementTree = None
         self.gpx_root: ET.Element = None
@@ -39,11 +43,39 @@ def __init__(self, file_path: str = "") -> None:
 
         self.gpx: Gpx = Gpx()
 
-        if self.file_path != "":
+        if self.file_path is not None:
             self.parse()
 
-    def check_schema(self):
-        pass
+    def check_schema(self, extensions_schema: bool = False) -> bool:
+        """
+        Check XML schema.
+
+        Args:
+            extensions_schema (bool, optional): Toogle extensions schema verificaton. Requires internet connection and is not guaranted to work.Defaults to False.
+
+        Returns:
+            bool: True if the file follows XML schemas.
+        """
+        if extensions_schema:
+            gpx_schemas = [s for s in self.gpx.xsi_schema_location if s.endswith(".xsd")]
+            for gpx_schema in gpx_schemas:
+                print(f"schema = {gpx_schema}")
+                schema = xmlschema.XMLSchema(gpx_schema)
+                if not schema.is_valid(self.file_path):
+                    logging.error(f"File does not follow {gpx_schema}")
+                    return False
+        else:
+            schema = None
+            if self.gpx.version == "1.1":
+                schema = xmlschema.XMLSchema("schemas/gpx_1_1/gpx.xsd")
+            elif self.gpx.version == "1.0":
+                schema = xmlschema.XMLSchema("schemas/gpx_1_0/gpx.xsd")
+
+            if schema is not None:
+                return schema.is_valid(self.file_path)
+            else:
+                logging.error("Unable to check XML schema")
+                return True
 
     def find_precision(self, number: str) -> int:
         """
@@ -623,18 +655,20 @@ def parse_root_extensions(self):
         extensions = self.gpx_root.find("topo:extensions", self.name_space)
         self.gpx.extensions = self.parse_extensions(extensions)
 
-    def parse(self, file_path: str = "") -> Gpx:
+    def parse(self, file_path: Optional[str] = None, check_schema: bool = True, extensions_schema: bool = False) -> Gpx:
         """
         Parse GPX file.
 
         Args:
-            file_path (str, optional): Path to the file to parse. Defaults to "".
+            file_path (str, optional): Path to the file to parse. Defaults to None.
+            check_schema (bool, optional): Toogle schema verificaton. Defaults to True.
+            extensions_schema (bool, optional): Toogle extensions schema verificaton. Requires internet connection and is not guaranted to work.Defaults to False.
 
         Returns:
             Gpx: Gpx instance., self.name_space).text
         """
         # File
-        if file_path != "":
+        if file_path is not None and os.path.exists(file_path):
             self.file_path = file_path
         elif self.file_path == "":
             logging.error("No GPX file to parse.")
@@ -655,6 +689,12 @@ def parse(self, file_path: str = "") -> Gpx:
             logging.error("Unable to parse properties in GPX file.")
             raise
 
+        # Check XML schema
+        if check_schema:
+            if not self.check_schema(extensions_schema):
+                logging.error("Invalid GPX file (does not follow XML schema).")
+                raise
+
         # Find precisions
         self.find_precisions()
 

diff --git a/ezgpx/schemas/gpx_1_0/gpx.xsd b/ezgpx/schemas/gpx_1_0/gpx.xsd
@@ -0,0 +1,231 @@
+<?xml version="1.0" encoding="utf-8"?>
+<!-- GPX.xsd version 1.0 - For more information on GPX and this schema, visit http://www.topografix.com/gpx.asp -->
+<xsd:schema
+	xmlns:xsd="http://www.w3.org/2001/XMLSchema"
+	xmlns:gpx="http://www.topografix.com/GPX/1/0"
+	targetNamespace="http://www.topografix.com/GPX/1/0"
+	elementFormDefault="qualified">
+
+<!-- Main GPX definition -->
+
+  <xsd:element name="gpx">
+    <xsd:complexType>
+      <xsd:sequence>
+        <xsd:element name="name"		type="xsd:string" minOccurs="0"/>		<!-- GPX file name -->
+        <xsd:element name="desc"		type="xsd:string" minOccurs="0"/>		<!-- GPX file description -->
+        <xsd:element name="author"		type="xsd:string" minOccurs="0"/>		<!-- GPX file author -->
+        <xsd:element name="email"		type="gpx:emailType" minOccurs="0"/>	<!-- GPX file author email -->
+        <xsd:element name="url"			type="xsd:anyURI" minOccurs="0"/>		<!-- GPX file URL -->
+ 		<xsd:element name="urlname"		type="xsd:string" minOccurs="0"/>
+        <xsd:element name="time"		type="xsd:dateTime" minOccurs="0"/>		<!-- GPX file creation time -->
+        <xsd:element name="keywords"	type="xsd:string" minOccurs="0"/>		<!-- GPX file keywords -->
+        <xsd:element name="bounds"		type="gpx:boundsType" minOccurs="0"/>	<!-- GPX file bounding rect -->
+        <xsd:element name="wpt" minOccurs="0" maxOccurs="unbounded">
+          <xsd:complexType>
+            <xsd:sequence>	<!-- elements must appear in this order -->
+			  <!-- Position info -->
+              <xsd:element name="ele"			type="xsd:decimal"		minOccurs="0"/>
+              <xsd:element name="time"			type="xsd:dateTime"		minOccurs="0"/>
+			  <xsd:element name="magvar"		type="gpx:degreesType"	minOccurs="0"/>
+			  <xsd:element name="geoidheight"	type="xsd:decimal"		minOccurs="0"/>
+
+			  <!-- Description info -->
+			  <xsd:element name="name"			type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="cmt"			type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="desc"			type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="src"			type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="url"			type="xsd:anyURI"		minOccurs="0"/>
+			  <xsd:element name="urlname"		type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="sym"			type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="type"			type="xsd:string"		minOccurs="0"/>
+
+			  <!-- Accuracy info -->
+			  <xsd:element name="fix"			type="gpx:fixType"		minOccurs="0"/>
+			  <xsd:element name="sat"			type="xsd:nonNegativeInteger"	minOccurs="0"/>
+			  <xsd:element name="hdop"			type="xsd:decimal"		minOccurs="0"/>
+			  <xsd:element name="vdop"			type="xsd:decimal"		minOccurs="0"/>
+			  <xsd:element name="pdop"			type="xsd:decimal"		minOccurs="0"/>
+			  <xsd:element name="ageofdgpsdata"	type="xsd:decimal"		minOccurs="0"/>
+			  <xsd:element name="dgpsid"		type="gpx:dgpsStationType"	minOccurs="0"/>
+
+			  <!-- you can add your own privately defined wpt elements at the end of the wpt -->
+              <xsd:any namespace="##other" minOccurs="0" maxOccurs="unbounded"/>
+            </xsd:sequence>
+            <xsd:attribute name="lat"			type="gpx:latitudeType"		use="required"/>
+            <xsd:attribute name="lon"			type="gpx:longitudeType"	use="required"/>
+          </xsd:complexType>
+        </xsd:element>
+        <xsd:element name="rte" minOccurs="0" maxOccurs="unbounded">
+          <xsd:complexType>
+            <xsd:sequence>
+              <xsd:element name="name"			type="xsd:string"	minOccurs="0"/>
+			  <xsd:element name="cmt"			type="xsd:string"	minOccurs="0"/>
+              <xsd:element name="desc"			type="xsd:string"	minOccurs="0"/>
+			  <xsd:element name="src"			type="xsd:string"	minOccurs="0"/>	<!-- the source of this data: "Garmin eTrex", "Map", etc -->
+			  <xsd:element name="url"			type="xsd:anyURI"	minOccurs="0"/>
+ 			  <xsd:element name="urlname"		type="xsd:string"	minOccurs="0"/>
+			  <xsd:element name="number"		type="xsd:nonNegativeInteger"	minOccurs="0"/>	<!-- GPS track number -->
+	<!--	  <xsd:element name="type"			type="xsd:string"	minOccurs="0"/>	PROPOSED -->
+			  <!-- you can add your own privately defined rte elements at the end of the rte -->
+              <xsd:any namespace="##other" minOccurs="0" maxOccurs="unbounded"/>
+              <xsd:element name="rtept"	minOccurs="0" maxOccurs="unbounded">
+                <xsd:complexType>
+                  <xsd:sequence>	<!-- elements must appear in this order -->
+
+				    <!-- Position info -->
+                    <xsd:element name="ele"		type="xsd:decimal"		minOccurs="0"/>
+                    <xsd:element name="time"	type="xsd:dateTime"		minOccurs="0"/>
+					<xsd:element name="magvar"	type="gpx:degreesType"	minOccurs="0"/>
+					<xsd:element name="geoidheight"	type="xsd:decimal"	minOccurs="0"/>
+
+					<!-- Description info -->
+                    <xsd:element name="name"	type="xsd:string"		minOccurs="0"/>
+                    <xsd:element name="cmt"		type="xsd:string"		minOccurs="0"/>
+                    <xsd:element name="desc"	type="xsd:string"		minOccurs="0"/>
+                    <xsd:element name="src"		type="xsd:string"		minOccurs="0"/>
+					<xsd:element name="url"		type="xsd:anyURI"		minOccurs="0"/>
+ 				    <xsd:element name="urlname"	type="xsd:string"		minOccurs="0"/>
+                    <xsd:element name="sym"		type="xsd:string"		minOccurs="0"/>
+                    <xsd:element name="type"	type="xsd:string"		minOccurs="0"/>
+
+					<!-- Accuracy info -->
+					<xsd:element name="fix"		type="gpx:fixType"		minOccurs="0"/>
+					<xsd:element name="sat"		type="xsd:nonNegativeInteger"	minOccurs="0"/>
+					<xsd:element name="hdop"	type="xsd:decimal"		minOccurs="0"/>
+					<xsd:element name="vdop"	type="xsd:decimal"		minOccurs="0"/>
+					<xsd:element name="pdop"	type="xsd:decimal"		minOccurs="0"/>
+					<xsd:element name="ageofdgpsdata"	type="xsd:decimal"	minOccurs="0"/>
+					<xsd:element name="dgpsid"	type="gpx:dgpsStationType"	minOccurs="0"/>
+
+					<!-- you can add your own privately defined rtept elements at the end of the rtept -->
+                    <xsd:any namespace="##other" minOccurs="0" maxOccurs="unbounded"/>
+                  </xsd:sequence>
+                  <xsd:attribute name="lat"		type="gpx:latitudeType"		use="required"/>
+                  <xsd:attribute name="lon"		type="gpx:longitudeType"	use="required"/>
+                </xsd:complexType>
+              </xsd:element>
+            </xsd:sequence>
+          </xsd:complexType>
+        </xsd:element>
+        <xsd:element name="trk" minOccurs="0" maxOccurs="unbounded">
+          <xsd:complexType>
+            <xsd:sequence>
+              <xsd:element name="name"			type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="cmt"			type="xsd:string"		minOccurs="0"/>
+              <xsd:element name="desc"			type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="src"			type="xsd:string"		minOccurs="0"/>	<!-- the source of this data: "Garmin eTrex", "Map", etc -->
+			  <xsd:element name="url"			type="xsd:anyURI"		minOccurs="0"/>
+ 			  <xsd:element name="urlname"		type="xsd:string"		minOccurs="0"/>
+			  <xsd:element name="number"		type="xsd:nonNegativeInteger"	minOccurs="0"/>	<!-- GPS track number -->
+	<!-- 	  <xsd:element name="type"			type="xsd:string"		minOccurs="0"/>  PROPOSED -->
+			  <!-- you can add your own privately defined trk elements at the end of the trk -->
+              <xsd:any namespace="##other" minOccurs="0" maxOccurs="unbounded"/>
+              <xsd:element name="trkseg"	minOccurs="0" maxOccurs="unbounded">
+                <xsd:complexType>
+                  <xsd:sequence>	<!-- elements must appear in this order -->
+				    <xsd:element name="trkpt"	minOccurs="0" maxOccurs="unbounded">
+					  <xsd:complexType>
+					    <xsd:sequence>	<!-- elements must appear in this order -->
+
+						  <!-- Position info -->
+						  <xsd:element name="ele"			type="xsd:decimal"		minOccurs="0"/>
+						  <xsd:element name="time"			type="xsd:dateTime"		minOccurs="0"/>
+						  <xsd:element name="course"		type="gpx:degreesType"	minOccurs="0"/>
+						  <xsd:element name="speed"			type="xsd:decimal"		minOccurs="0"/>
+						  <xsd:element name="magvar"		type="gpx:degreesType"	minOccurs="0"/>
+						  <xsd:element name="geoidheight"	type="xsd:decimal"		minOccurs="0"/>
+
+						  <!-- Description info -->
+						  <xsd:element name="name"			type="xsd:string"		minOccurs="0"/>
+						  <xsd:element name="cmt"			type="xsd:string"		minOccurs="0"/>
+						  <xsd:element name="desc"			type="xsd:string"		minOccurs="0"/>
+						  <xsd:element name="src"			type="xsd:string"		minOccurs="0"/>
+						  <xsd:element name="url"			type="xsd:anyURI"		minOccurs="0"/>
+ 						  <xsd:element name="urlname"		type="xsd:string"		minOccurs="0"/>
+						  <xsd:element name="sym"			type="xsd:string"		minOccurs="0"/>
+						  <xsd:element name="type"			type="xsd:string"		minOccurs="0"/>
+
+						  <!-- Accuracy info -->
+						  <xsd:element name="fix"			type="gpx:fixType"		minOccurs="0"/>
+						  <xsd:element name="sat"			type="xsd:nonNegativeInteger"	minOccurs="0"/>
+						  <xsd:element name="hdop"			type="xsd:decimal"		minOccurs="0"/>
+						  <xsd:element name="vdop"			type="xsd:decimal"		minOccurs="0"/>
+						  <xsd:element name="pdop"			type="xsd:decimal"		minOccurs="0"/>
+						  <xsd:element name="ageofdgpsdata"	type="xsd:decimal"		minOccurs="0"/>
+						  <xsd:element name="dgpsid"		type="gpx:dgpsStationType"		minOccurs="0"/>
+
+						  <!-- you can add your own privately defined trkpt elements at the end of the trkpt -->
+						  <xsd:any namespace="##other" minOccurs="0" maxOccurs="unbounded"/>
+					    </xsd:sequence>
+					    <xsd:attribute name="lat"			type="gpx:latitudeType"		use="required"/>
+					    <xsd:attribute name="lon"			type="gpx:longitudeType"	use="required"/>
+					  </xsd:complexType>
+				    </xsd:element>
+			  	  </xsd:sequence>
+			    </xsd:complexType>
+			  </xsd:element>
+            </xsd:sequence>
+          </xsd:complexType>
+        </xsd:element>
+		<!-- you can add your own privately defined elements at the end of the GPX file -->
+        <xsd:any namespace="##other" minOccurs="0" maxOccurs="unbounded"/>
+      </xsd:sequence>
+      <xsd:attribute name="version" type="xsd:string" use="required" fixed="1.0"/>	<!-- version 1.0 -->
+      <xsd:attribute name="creator" type="xsd:string" use="required"/>
+    </xsd:complexType>
+  </xsd:element>
+
+  <!-- Other types used by GPX -->
+
+  <xsd:simpleType name="latitudeType">
+    <xsd:restriction base="xsd:decimal">
+      <xsd:minInclusive value="-90.0"/>
+      <xsd:maxInclusive value="90.0"/>
+    </xsd:restriction>
+  </xsd:simpleType>
+
+  <xsd:simpleType name="longitudeType">
+    <xsd:restriction base="xsd:decimal">
+      <xsd:minInclusive value="-180.0"/>
+      <xsd:maxInclusive value="180.0"/>
+    </xsd:restriction>
+  </xsd:simpleType>
+
+  <xsd:simpleType name="degreesType">	<!-- for bearing, heading, course.  Units are degrees, true -->
+    <xsd:restriction base="xsd:decimal">
+      <xsd:minInclusive value="0.0"/>
+      <xsd:maxInclusive value="360.0"/>
+    </xsd:restriction>
+  </xsd:simpleType>
+
+  <xsd:simpleType name="fixType">
+    <xsd:restriction base="xsd:string">
+      <xsd:enumeration value="none"/>	<!-- none means GPS had no fix.  To signify "the fix info is unknown, leave out the <fix> tag entirely --> 
+      <xsd:enumeration value="2d"/>
+      <xsd:enumeration value="3d"/>
+      <xsd:enumeration value="dgps"/>
+      <xsd:enumeration value="pps"/>		<!-- military signal used -->
+    </xsd:restriction>
+  </xsd:simpleType>
+
+  <xsd:simpleType name="dgpsStationType">
+    <xsd:restriction base="xsd:integer">
+      <xsd:minInclusive value="0"/>
+      <xsd:maxInclusive value="1023"/>
+    </xsd:restriction>
+  </xsd:simpleType>
+
+  <xsd:complexType name="boundsType">	<!-- bounding rect for data in file -->
+    <xsd:attribute name="minlat"		type="gpx:latitudeType"		use="required"/>
+    <xsd:attribute name="minlon"		type="gpx:longitudeType"	use="required"/>
+    <xsd:attribute name="maxlat"		type="gpx:latitudeType"		use="required"/>
+    <xsd:attribute name="maxlon"		type="gpx:longitudeType"	use="required"/>
+  </xsd:complexType>
+
+  <xsd:simpleType name="emailType">
+    <xsd:restriction base="xsd:string">
+      <xsd:pattern value="[\p{L}_]+(\.[\p{L}_]+)*@[\p{L}_]+(\.[\p{L}_]+)+"/>
+    </xsd:restriction>
+  </xsd:simpleType>
+
+</xsd:schema>