UltraEdit filters out all results of the regular expression finds which have as result an empty string because of listing multiple empty strings in function list view is not really helpful for any programmer. That would be just confusing in my opinion, at least I would be really confused on seeing empty lines at top of a sorted or somewhere in the middle of a non-sorted function list.
It makes sense to display in function list just the function name and the function parameters for functions with a name. My suggestion for anonymous functions is displaying everything after leading tabs/spaces to end of the keyword
function and the parameters of the anonymous functions.
Please open the wordfile used to syntax highlight JavaScript code. This can be done by opening in UltraEdit a JavaScript file like the file with the example lines, next open
Advanced - Settings or Configuration - Editor display - Syntax highlighting, then click on button
Open right to the already selected installed wordfile for JavaScript syntax highlighting, and finally close the configuration dialog with a click on button
Cancel or symbol
X or with key
ESC. These steps opens with default configuration the wordfile
%APPDATA%\IDMComp\UltraEdit\wordfiles\javascript.uew.
The wordfile
javascript.uew installed with currently latest UltraEdit for Windows v27.00.0.72 contains at top the following block:
Code: Select all
/TGBegin "Functions"
/TGFindStr = "^[ \t]*function[ \t]*(\w+[ \t]*)\(.*\)"
/TGFindStr = "^[ \t]*function[ \t]*(\w+[ \t]*)\([^\)]*?\)"
/TGBegin "Parameters"
/TGFindStr = "[ \t\n]*([^,\s]+)"
/TGFindBStart = "\("
/TGFindBEnd = "\)"
/TGEnd
/TGEnd
/TGBegin "Anonymous Functions"
/TGFindStr = "(?:var[ \t]+)?((?:[\w\.]+))[ \t]*?[:=][ \t]*?function[ \t]*?\([^\)]*?\)"
/TGEnd
This block must be replaced by the following block:
Code: Select all
/TGBegin "Functions"
/TGFindStr = "(?-i)^[\t ]*(?:[!(]|\w+\(|(?:(?:const|let|var)[\t ]+)?[\w$]+[\t ]*=[\t ]*(?:{\w+[\t ]*:[\t ]*)?)?function[\t ]*([\w$]+)[\t ]*\([^\)]*?\)"
/TGBegin "Parameters"
/TGFindStr = "[\t\n ]*([^,\s]+)"
/TGFindBStart = "(?-i)function[\t ]*[\w$]+[\t ]*\("
/TGFindBEnd = "\)"
/TGEnd
/TGEnd
/TGBegin "Anonymous Functions"
/TGFindStr = "(?-i)^[\t ]*((?:[!(]|\w+\(|(?:(?:const|let|var)[\t ]+)?[\w$]+[\t ]*=[\t ]*(?:{\w+[\t ]*:[\t ]*)?)?function)[\t ]*[\t ]*\([^\)]*?\)"
/TGBegin "Parameters"
/TGFindStr = "[\t\n ]*([^,\s]+)"
/TGFindBStart = "(?-i)function[\t ]*\("
/TGFindBEnd = "\)"
/TGEnd
/TGEnd
UltraEdit runs the regular expression finds defined in a syntax highlighting wordfile always case-insensitive. I used the Perl regular expression flag
(?-i) at beginning of the search expression strings containing JavaScript keywords like
function,
const,
let, or
var to run the finds case-sensitive for avoiding false positives.
I added non-word character
$ as this is a valid character in an identifier name in addition to the letters according to the
Draft ECMA-262 / July 7, 2020 of ECMAScript® 2021 Language Specification.
I haven't added zero width non-joiner (U+200C) and zero width joiner (U+200D) which would be valid characters in an identifier name on being not the first character as this would make the Perl regular expression search strings invalid expressions on being executed on a non-Unicode encoded JavaScript file with older versions of UltraEdit.
The expressions above were tested by me on your example with some additions with UltraEdit for Windows v22.20 and v27.00.
It would be also possible to use the following block:
Code: Select all
/TGBegin "Functions"
/TGFindStr = "(?-i)^.*\bfunction[\t ]*([\w$]+)[\t ]*\([^\)]*?\)"
/TGBegin "Parameters"
/TGFindStr = "[\t\n ]*([^,\s]+)"
/TGFindBStart = "(?-i)function[\t ]*[\w$]+[\t ]*\("
/TGFindBEnd = "\)"
/TGEnd
/TGEnd
/TGBegin "Anonymous Functions"
/TGFindStr = "(?-i)^[\t ]*(.*\bfunction)[\t ]*[\t ]*\([^\)]*?\)"
/TGBegin "Parameters"
/TGFindStr = "[\t\n ]*([^,\s]+)"
/TGFindBStart = "(?-i)function[\t ]*\("
/TGFindBEnd = "\)"
/TGEnd
/TGEnd
The two main case-sensitive interpreted Perl regular expression search strings are less restrictive as it can be seen quite easily. But it could happen to get false positives on using these simplified search expressions.
Note: A user has the option to get the functions listed as found in file or sorted
alphabetically via
Function List context menu option
Sort List. The functions are not sorted
alphanumeric. Meaningful function names do not usually start all with same string and the difference is just a number at end of function name with individual number of digits. I have definitely no source file on which an entire function name is the start of another function name. I avoid in all programing and scripting languages identifier names which are included completely in another identifier name. I have learned what is a good and what is a not so good naming scheme in more than 25 years of programming when I need to search for an identifier name easily in thousands of source code files. I don't like it to be forced to use a regular expression find to find one specific identifier name in dozens, hundreds or even thousands of source code files.
