The main problem is the line
Code: Select all
var Regex = /<clstep.*?>.*?<\/clstep1>/img;
The dot does not match the newline characters carriage return and line-feed. For that reason this JavaScript regular expression object on execution of a procedure block string does not find anything to match. One solution would be using the flag
s as documented on
RegExp documentation page. But it depends on version of JavaScript core and therefore on version of UltraEdit if this flag is supported at all.
The flag
m is counter-productive for this regular expression search and the flag
i is not needed here because of XML files are case-sensitive and for that reason a slower case-insensitive search should be never necessary on processing an XML file with a script or macro using finds/replaces searching for the case-sensitive tags in XML file.
The solution working with any version of JavaScript core engine is:
Code: Select all
var Regex = /<clstep.*?>[\s\S]*?<\/clstep\d+>/g;
The character class definition
[\s\S] matches any whitespace character (including carriage return and line-feed) or any non-whitespace character, or in other words really any character. The regular expression object with this regular expression search string and flag
g should be defined once outside of all loops for best efficiency.
Another mistake is the fact that method
exec() returns not a single string, but an array of strings, or more precise an array of a structure with first element being the found string, element
index being the character index in input string on which the found string begins and element
input being the entire input string searched by the method. This can be seen on using function
var_dump which is a function added by UltraEdit and UEStudio to every script executed by UE/UES in the script file created in temporary files directory with all other script files included on using the special
// include directive before the script is really executed by built-in JavaScript core engine.
Here is your script with the small changes as described above.
Code: Select all
function XMLConv() {
//Define environment for this script.
if (UltraEdit.document.length > 0) // Is any file opened?
{
// Define environment for this script.
UltraEdit.insertMode();
if (typeof(UltraEdit.columnModeOff) == "function") UltraEdit.columnModeOff();
else if (typeof(UltraEdit.activeDocument.columnModeOff) == "function") UltraEdit.activeDocument.columnModeOff();
UltraEdit.selectClipboard(8);
// Get document index number
var nActiveFileIndex = UltraEdit.activeDocumentIdx;
UltraEdit.perlReOn();
UltraEdit.document[nActiveFileIndex].top();
UltraEdit.document[nActiveFileIndex].findReplace.mode = 0;
UltraEdit.document[nActiveFileIndex].findReplace.matchCase = true;
UltraEdit.document[nActiveFileIndex].findReplace.matchWord = false;
UltraEdit.document[nActiveFileIndex].findReplace.regExp = true;
UltraEdit.document[nActiveFileIndex].findReplace.searchDown = true;
UltraEdit.document[nActiveFileIndex].findReplace.searchInColumn = false;
var Regex = /<clstep.*?>[\s\S]*?<\/clstep1>/g;
UltraEdit.document[nActiveFileIndex].top();
while (UltraEdit.document[nActiveFileIndex].findReplace.find("<procedure")) {
// Remember in two variables the current caret position of
// the chapter beginning in file. There is nothing selected!
var nChLine = UltraEdit.document[nActiveFileIndex].currentLineNum;
//UltraEdit.activeDocument.key("HOME");
var nChColumn = UltraEdit.document[nActiveFileIndex].currentColumnNum;
if (UltraEdit.document[nActiveFileIndex].findReplace.find("</procedure>")) {
UltraEdit.activeDocument.key("RIGHT ARROW");
UltraEdit.document[nActiveFileIndex].gotoLineSelect(nChLine, nChColumn);
var procedure = UltraEdit.document[nActiveFileIndex].selection;
var clStep = Regex.exec(procedure);
if (clStep) var_dump(clStep);
}
}
}
}
XMLConv();
However, on processing a small XML file with less than 20 MB it would be most likely better to load entire file content into memory and process the XML file in memory, and on making changes, write the processed data back to file overwriting entire file content. That is much faster as it avoids window updates and produces only one undo step to restore the file content before script execution.
Here is the code for this alternate method to process a small XML file:
Code: Select all
function XMLConv() {
// Define environment for this script.
UltraEdit.insertMode();
if (typeof(UltraEdit.columnModeOff) == "function") UltraEdit.columnModeOff();
else if (typeof(UltraEdit.activeDocument.columnModeOff) == "function") UltraEdit.activeDocument.columnModeOff();
// Select entire file content.
UltraEdit.activeDocument.selectAll();
// Get the file content as array of strings with splitting the
// file on each occurrence of end tag of XML element procedure.
var asProcedures = UltraEdit.activeDocument.selection.split("</procedure>");
if (!asProcedures)
{
UltraEdit.activeDocument.top();
return;
}
var rStep = /<clstep.*?>[\s\S]*?<\/clstep[0-9]+>/g;
// The last string in array must be always ignored.
var nProcedureCount = asProcedures.length - 1;
for (var nProcedureIndex = 0; nProcedureIndex < nProcedureCount; nProcedureIndex++)
{
var asStep = asProcedures[nProcedureIndex].match(rStep);
if (asStep)
{
for (var nStepIndex = 0; nStepIndex < asStep.length; nStepIndex++)
{
UltraEdit.outputWindow.write("Procedure " + nProcedureIndex +
", step " + nStepIndex + ":\n" +
asStep[nStepIndex]);
}
}
// Append the end tag removed on file content splitting.
asProcedures[nProcedureIndex] += "</procedure>";
}
// UltraEdit.activeDocument.write(asProcedures.join(""));
}
if (UltraEdit.document.length > 0) XMLConv();
The entire file content is selected and split into substrings on each occurrence of string
</procedure>. The last string in array of strings is the file content after last
</procedure> which can be also an empty string. This last string in array must be ignored for that reason.
There is again a regular expression search executed on each procedure block to get the
clstep1 elements because I am not sure if the number is really fixed in real XML file, i.e. there are
clstep1,
clstep2, etc. within a
procedure element. The string method
split() with string
</clstep1> would be otherwise more efficient to get an array of
clstep1 elements (with procedure data on first element and last string being ignored).
On each
procedure block is missing
</procedure> at end because of string method
split() removes this string. This must be taken into account on finally joining the procedure blocks again and writing entire file content back to the file overwriting still selected entire file content. In the example code a currently commented line would append
</procedure> to each procedure block.
Attention: The simple usage of string method
split() on processing an XML file can be used only if no element can exist nested on which split on end tag is done. In other words the splitting method cannot be used if a
procedure element contains itself a
procedure element or a
clstep contains itself a
clstep element.
UltraEdit has an XML parser built-in used for
XML Manager view. But the DOM (document object model) of the XML file is not accessible from within the scripting environment in UltraEdit for Windows v26.20.0.68 and all former versions to parse the nodes and modify the node objects to reformat an XML file based on its XML nodes by a script.
