Regex with PowerShell won't work with /n (newline)

Question

For some reason, newline characters will not work with my regex.

Here's the first part of the PowerShell script. This piece of code is used to get a set of markdown files. For testing purposes, I'm only using the first file.

The code then gets the content inside the file.

$path = 'C:\Users\Will\Desktop\ProjectTemp'

$filelist = ls $path -filter *.md -recurse

[string[]]$currentfile_path = $filelist[0].FullName

$currentfile_data = Get-Content $currentfile_path

This is the problematic code. It works fine anywhere else, but PowerShell will not find the matched so that they can be replaced. Specifically, I have found that the newline characters will not work. Nothing I have tried will work.

$currentfile_data -Replace '([ \t]*)\| (.*) \| (.*) \| (.*) \|\r\n[ \t]*\| .* \| .* \| .* \|\r\n[ \t]*\| (.*) \| (.*) \| (.*) \|','$1<table>\n$1\t<tr align="center" valign="middle">\n$1\t\t<th>$2</th>\n$1\t\t<th>$3</th>\n$1\t\t<th>$4</th>\n$1\t</tr>\n$1\t<tr align="center" valign="middle">\n$1\t\t<td>$5</td>\n$1\t\t<td>$6</td>\n$1\t\t<td>$7</td>\n$1\t</tr>\n$1</table>'

This is meant to find the regex: '([ \t]*)\| (.*) \| (.*) \| (.*) \|\r\n[ \t]*\| .* \| .* \| .* \|\r\n[ \t]*\| (.*) \| (.*) \| (.*) \|'

and replace it with:

$1<table>\n$1\t<tr align="center" valign="middle">\n$1\t\t<th>$2</th>\n$1\t\t<th>$3</th>\n$1\t\t<th>$4</th>\n$1\t</tr>\n$1\t<tr align="center" valign="middle">\n$1\t\t<td>$5</td>\n$1\t\t<td>$6</td>\n$1\t\t<td>$7</td>\n$1\t</tr>\n$1</table>'

What the input and output should look like:

| th1 | th2 | th3 |
| :-: | :-: | :-: |
| td1 | td2 | td3 |

<table>
    <tr align="center" valign="middle">
        <th>th1</th>
        <th>th2</th>
        <th>th3</th>
    </tr>
    <tr align="center" valign="middle">
        <td>td1</td>
        <td>td2</td>
        <td>td3</td>
    </tr>
</table>

Is there any way this can be fixed? Am I doing something wrong?

Answer 1

tl;dr

$currentfile_data = Get-Content -Raw $currentfile_path

-Raw ensures that Get-Content reads the iput file as a whole, as a single, typically multi-line string, which ensures that the subsequent -replace operation can perform matching across lines.

---

As for what you tried:

$currentfile_data = Get-Content $currentfile_path

This saves an array of the lines from the input file in variable $currentfile_data, because Get-Content by default streams the target file's lines one by one.

$currentfile_data -Replace '...'

This makes the -replace operator act on each element of the array stored in $currentfile_data, i.e. on each line of the original file rather than across lines.

---

A PowerShell (Core) 7+ alternative:

As Santiago Squarzon points out, PowerShell (Core) 7+ ships with the ConvertFrom-Markdown cmdlet, which can directly transform your input file to HTML (albeit not in the exact same format):

(@'
| th1 | th2 | th3 |
| :-: | :-: | :-: |
| td1 | td2 | td3 |
'@ | ConvertFrom-Markdown).Html

Output:

<table>
<thead>
<tr>
<th style="text-align: center;">th1</th>
<th style="text-align: center;">th2</th>
<th style="text-align: center;">th3</th>
</tr>
</thead>
<tbody>
<tr>
<td style="text-align: center;">td1</td>
<td style="text-align: center;">td2</td>
<td style="text-align: center;">td3</td>
</tr>
</tbody>
</table>