Document Toolkit/v0.1

Document & File Plumbing Toolkit

Copy-pasteable expression rules for the document and file pain points Appian developers hit every week. Forum-signal driven, shipped as ready-to-deploy snippets with the explainer paragraph you wish was on the docs page.

2/2

DOCXPDFSmart Service

Preserve line breaks in DOCX-to-PDF

PDF From DOCX smart service flattens char(10) line breaks in dynamic paragraph text. Merged output reads like one long sentence.

a!localVariables(
  local!raw: ri!userInputText,
  local!pieces: split(local!raw, char(10)),
  local!withBreaks: joinarray(
    local!pieces,
    "</w:t><w:br/><w:t xml:space=""preserve"">"
  ),
  local!docxSafe: "<w:t xml:space=""preserve"">" & local!withBreaks & "</w:t>",
  local!docxSafe
)

How it works

The merge engine strips raw newlines before the renderer ever sees them, so the fix is to rewrite the input as DOCX-native run XML before binding it to the placeholder. Split on char(10), wrap each fragment in a <w:t> run, and stitch them back together with <w:br/> in between. xml:space="preserve" is non-optional - Word will trim leading whitespace on every fragment without it, and the symptoms look identical to the original bug. Bind this rule output directly to the placeholder, not the raw input.

PDFDocument ExtractionPre-condition

Detect an encrypted PDF before doc-extract

There is no native isEncrypted function in Appian, so encrypted PDFs slip through and the document-extraction smart service fails silently with an empty result.

a!localVariables(
  local!header: tostring(
    a!fileBytes(ri!doc, 0, 1024)
  ),
  local!hasEncryptDict: or(
    contains(local!header, "/Encrypt "),
    contains(local!header, "/Encrypt<<")
  ),
  local!hasFilterStandard: contains(local!header, "/Filter/Standard"),
  if(
    or(local!hasEncryptDict, local!hasFilterStandard),
    "ENCRYPTED",
    "OPEN"
  )
)

How it works

Every encrypted PDF stamps an /Encrypt entry in the trailer or xref. In the overwhelming majority of real-world docs that marker sits inside the first 1024 bytes, so a!fileBytes lets you peek at the header without loading the full file into memory. /Encrypt has two common renderings depending on the producer (with a trailing space, or directly followed by <<) and /Filter/Standard catches the rarer streamed-out case. Run this rule as a precondition on the doc-extract sub-process - reject and route to a manual queue if ENCRYPTED, otherwise pass it through. Cheap, runs in milliseconds, kills the silent-failure mode.

Coming next

Hyperlink-preserving Excel export, multi-source doc-gen template recipe, Excel Tools migration recipes, one-click XML/snippet export, and tag/filter UI by file type.

a!localVariables( local!raw: ri!userInputText, local!pieces: split(local!raw, char(10)), local!withBreaks: joinarray( local!pieces, "</w:t><w:br/><w:t xml:space=""preserve"">" ), local!docxSafe: "<w:t xml:space=""preserve"">" & local!withBreaks & "</w:t>", local!docxSafe )

a!localVariables( local!header: tostring( a!fileBytes(ri!doc, 0, 1024) ), local!hasEncryptDict: or( contains(local!header, "/Encrypt "), contains(local!header, "/Encrypt<<") ), local!hasFilterStandard: contains(local!header, "/Filter/Standard"), if( or(local!hasEncryptDict, local!hasFilterStandard), "ENCRYPTED", "OPEN" ) )