fix PPR condor correction to keep correct format and encoding

apps/cli/executables/pexable/casa_envoy/casa_envoy/auditor.py

@@ -25,13 +25,34 @@ import shutil
 from pathlib import Path
 from typing import Dict, List, Union
 
-import prettierfier
+import lxml.etree
 from bs4 import BeautifulSoup
+from lxml import etree
 
 from casa_envoy.enums import ProductType
 from casa_envoy.interfaces import AuditorIF
 from casa_envoy.schema import AbstractTextFile
 
+PPR_HEADER = """<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+"""
+PPR_FORMATTER = """<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
+    <xsl:output indent="yes"/>
+    <xsl:strip-space elements="*"/>
+
+    <!-- IDENTITY TRANSFORM -->
+    <xsl:template match="@*|node()">
+      <xsl:copy>
+        <xsl:apply-templates select="@*|node()"/>
+      </xsl:copy>
+    </xsl:template>
+
+    <!-- RUN normalize-space() ON ALL TEXT NODES -->
+    <xsl:template match="text()">
+        <xsl:copy-of select="normalize-space()"/>
+    </xsl:template>            
+</xsl:stylesheet>
+"""
+
 
 def get_fields_for(product_type: str, filename: str) -> list:
     cal_metadata_list = [
@@ -128,6 +149,12 @@ class AuditFiles(AuditorIF):
             return True
 
     def correct_for_condor(self, ppr: AbstractTextFile) -> AbstractTextFile:
+        """
+        Correct PPR.xml for HTCondor directory and path
+
+        :param ppr: content of the waiting PPR
+        :return: updated content of PPR written to working area
+        """
         shutil.copy(ppr.filename, "./working")
         os.chdir("./working")
 
@@ -135,8 +162,22 @@ class AuditFiles(AuditorIF):
         parsed_xml.find("RootDirectory").string = self.settings["rootDirectory"]
         parsed_xml.find("RelativePath").string = self.settings["processingDirectory"]
 
+        strip_declaration = str(parsed_xml).split("\n")[1:]
+        fixed_declaration = "".join(strip_declaration)
+
+        corrected_xml = etree.fromstring(fixed_declaration)
+        formatter = etree.XSLT(etree.fromstring(PPR_FORMATTER))
+
+        corrected_content = lxml.etree.tostring(
+            formatter(corrected_xml),
+            pretty_print=True,
+            xml_declaration=True,
+            encoding="UTF-8",
+            standalone=True,
+        )
+
         with open(ppr.filename, "w") as file:
-            file.write(prettierfier.prettify_xml(parsed_xml.prettify(), indent=4))
+            file.write(corrected_content.decode())
 
         os.chdir("../")
         return ppr
@@ -150,7 +191,9 @@ class AuditFiles(AuditorIF):
                 self.logger.info("Correcting PPR.xml for condor processing...")
                 file = self.correct_for_condor(file)
 
-            valid = self.check_required_fields(file=file, fields=get_fields_for(self.product_type, file.filename))
+            valid = self.check_required_fields(
+                file=file, fields=get_fields_for(self.product_type, file.filename)
+            )
             if not valid:
                 invalid_files.append(file.filename)
 
@@ -173,7 +216,9 @@ class AuditDirectories(AuditorIF):
         current = os.getcwd()
         needed = self.rootDirectory + "/" + self.relative_path
         if needed != current:
-            self.logger.error("DIRECTORY ERROR: not in correct directory for processing.")
+            self.logger.error(
+                "DIRECTORY ERROR: not in correct directory for processing."
+            )
             return False
         else:
             working = Path(current + "/working").is_dir()
@@ -190,9 +235,13 @@ class AuditDirectories(AuditorIF):
                         self.logger.info("Checking products/ for calibration tables...")
                         cal_data = os.listdir(Path(current + "/products/"))
                         if len(cal_data) > 0:
-                            self.logger.info("Calibration data is available. Proceeding...")
+                            self.logger.info(
+                                "Calibration data is available. Proceeding..."
+                            )
                         else:
-                            self.logger.error("FAILURE: calibration data not found in products/")
+                            self.logger.error(
+                                "FAILURE: calibration data not found in products/"
+                            )
                             return False
                     # if (
                     #     self.parameters["product_type"] == ProductType.VLASS_IMG.value
@@ -217,4 +266,6 @@ class AuditDirectories(AuditorIF):
                     self.logger.error("FAILURE: data not found in rawdata/")
                     return False
             else:
-                self.logger.error("DIRECTORY ERROR: A directory is missing from the processing root directory.")
+                self.logger.error(
+                    "DIRECTORY ERROR: A directory is missing from the processing root directory."
+                )

apps/cli/executables/pexable/casa_envoy/test/input_files/PPR.xml

-<?xml version="1.0" ?>
+<?xml version='1.0' encoding='UTF-8' standalone='yes'?>
 <ns2:SciPipeRequest xmlns:ns2="Common/pipelinescience/SciPipeRequest">
     <ProjectSummary>
         <ProposalCode>VLA/null</ProposalCode>