Create a text file mirror from PDFs, requires poppler

text-mirror.ps1

<#
    1. Leaf
    Given a text file of PDF filenames, extract content from PDFs recursively
    and create mirror directory structure for text file outputs.

    * Handles filenames with entry separators.
    * Ignores PDF older than its text file mirror unless -force param is used.
    * Requires poppler pdftotext.exe

    .\text-mirror.ps1 -in C:\dev\abc\extract\extracted\pdfs\abc-pdfs-1.txt `
        -pathprefix C:\dev\abc\src\web\media\documents `
        -outdir C:\dev\abc\raw\text-mirror\documents `
        -poppler C:\utils\poppler-0.68\bin\pdftotext.exe  

    2. Container
    Given a directory path, extract content from PDFs recursively
    and create mirror directory structure for text file outputs.

    * Ignores PDF older than its text file mirror unless -force param is used.
    * Requires poppler pdftotext.exe

    .\text-mirror.ps1 -indir C:\dev\abc\src\web\media\documents -outdir C:\dev\abc\raw\text-mirror\documents

    Peter Tyrrell, Andornot

#>

param (

    [Parameter(Mandatory=$true, ParameterSetName="Leaf")]
    [string]$in,

    [Parameter(Mandatory=$true, ParameterSetName="Leaf")]
    [string]$pathsnip,

    [Parameter(Mandatory=$true, ParameterSetName="Leaf")]
    [string]$pathprefix,

    [Parameter(Mandatory=$true, ParameterSetName="Container")]
    [string]$indir,

    [Parameter(Mandatory=$true, ParameterSetName="Leaf")]
    [Parameter(Mandatory=$false, ParameterSetName="Container")]
    [string]$outdir = $indir,

    [Parameter(Mandatory=$false, ParameterSetName="Leaf")]
    [Parameter(Mandatory=$false, ParameterSetName="Container")]
    [Switch]
    $force,

    [Parameter(Mandatory=$false, ParameterSetName="Leaf")]
    [Parameter(Mandatory=$false, ParameterSetName="Container")]
    [string]$logsrc = "",

    [Parameter(Mandatory=$false, ParameterSetName="Leaf")]
    [Parameter(Mandatory=$false, ParameterSetName="Container")]
    [string]$poppler = "c:\utils\poppler-0.68\bin\pdftotext.exe"
)

<#
    FUNCTIONS
#>

. .\helper.logging.ps1

function getValidPaths([string]$in, [string]$pathprefix, [string]$pathsnip) {

    $paths = get-content $in | Where-Object { $_ -ne '' } | ForEach-Object { $_.split('|') }

    $i = 0
    $e = 0
    $validPaths = @()
    foreach ($path in $paths) {
        $i++
        try {
            if ($pathsnip) {
                $path = $path.Replace($pathsnip, "", [System.StringComparison]::OrdinalIgnoreCase)
            }
            if ($pathprefix) {
                $path = join-path $pathprefix $path
            }
            $path = $path.trim()

            $ok = test-path $path -PathType Leaf -ErrorAction Stop
            if ($ok) {
                write-host "$path Found"
                $validPaths = $validPaths + $path
            }
            else {
                write-host "$path Not Found"
            }

            # update progress display
            write-progress -activity "Validating file paths..." -status "Processing $i of $($paths.Count)" -percentcomplete (($i / $paths.Count) * 100) 
        }
        catch [Exception] {
            $e++
            $msg = "Error occurred while validating $path. $($_.Exception)"
            logError $logsrc $msg
        }
    }
    return $validPaths
}

<# 
    MAIN
#>

if ($indir) {
    $files = get-childitem "$indir\*" -include *.pdf -recurse
    $pathprefix = $indir
}
if ($in) {
    $files = get-childitem $(getValidPaths $in $pathprefix $pathsnip) -include *.pdf
}

$i = 0
$e = 0
$skip = 0
foreach ($file in $files) {
  $i++
  $update = $true
  $status = "Processing $i of $($files.Count)."
  write-progress -activity "Extracting PDF text..." -status $status -percentcomplete (($i / $files.Count) * 100) 

  try {
    # txt file path => [base out dir] + [pdf parent dir w/o qualifier] + .txt
    $txtdir = join-path $outdir  $($file.DirectoryName.Replace($pathprefix, "", [System.StringComparison]::OrdinalIgnoreCase))
    $txtpath = [System.IO.Path]::Combine($txtdir, $("{0}.txt" -f $file.BaseName))

    # if txt already exists and timestamp is newer than pdf, skip (unless -force is active)
    if ((-not $force) -and (test-path $txtpath)) {
      $txtfile = get-itemproperty -path $txtpath
      if ($txtfile.LastWriteTime -gt $file.LastWriteTime) {
        $update = $false
        $status = $status + " (Skipping - $($file.FullName) doesn't need update.)"
      }
    }

    # extract text to txt
    if ($force -or $update) {
      if(!(test-path $txtdir)) {
        mkdir $txtdir | out-null
      }

      & $poppler -eol unix -raw -enc UTF-8 $file.FullName $txtpath

      # fix filenames mangled by poppler (writes filename diacritics with Windows-1252 instead of UTF-8)
      $txtpath_1252 = [System.Text.Encoding]::GetEncoding(1252).GetString([System.Text.Encoding]::UTF8.GetBytes($txtpath))
      if ($txtpath -ne $txtpath_1252) {
        move-item -path $txtpath_1252 -destination $txtpath
      }
    }
    else {
        $skip++
    }
  }
  catch [Exception] {
    $e++
    $msg = "Error occurred while processing $($file.FullName). $($_.Exception)"
    logError $logsrc $msg
  }
}

# final report
$msg = "PDF text extraction finished processing $i {0} to $outdir. $skip skipped, $e {1}." -f $(if ($e -eq 1) {"file"} else {"files"}),$(if ($e -eq 1) {"error"} else {"errors"})
logInfo $logsrc $msg

exit 0