PdfSharpDslCore

This is a sample library that uses Irony.Net to define a grammar and TerraPDF to print PDF.

Current support

PdfSharpDslCore multi-targets net8.0 and net10.0; PdfSharpDsl.Language and PdfSharpDslCore.Generator target netstandard2.0 because they are loaded by the compiler.
PdfSharpDslConsole and the test project target net10.0.
The repository is pinned to SDK 10.0.400 in global.json.
Version, licence, package metadata and every NuGet version are centralized in _build/: Version.props (the single product version), Common.props (shared metadata and target-framework aliases) and Packages.props (central package management). The Directory.Build.props, Directory.Build.targets and Directory.Packages.props files at the repository root only import these. Change a version or a target framework there, never in an individual .csproj.
The generator package includes its analyzer dependencies and supports clean NuGet consumer builds.

Run the full test and core coverage gate with:

powershell -NoProfile -File .\scripts\coverage.ps1

The gate requires at least 90% core line coverage. See tasks.md for the migration tracker and verified results.

Example

var parser = new Irony.Parsing.Parser(new PdfGrammar());

var parsingResult = parser.Parse(File.ReadAllText("pdfsharp.ipdf"));

if (parsingResult.HasErrors())
{
    //show Error
    foreach (var error in parsingResult.ParserMessages)
    {
        Console.Write(error.Location.ToString());
        Console.Write("=>");
        Console.WriteLine(error);
    }
}
else
{
    using var drawer = new PdfDocumentDrawer();
    new PdfDrawerVisitor().Draw(drawer, parsingResult);
    drawer.PublishPdf("helloworld.pdf");
}

or download source code, then goto PdfSharpDslConsole and run it

dotnet run

Architecture

sequenceDiagram
    participant yourprogram
    participant PdfSharpDslCore
    participant Irony
    participant TerraPDF
    yourprogram->>Irony: Parse file or text.
    Irony -->> PdfSharpDslCore: use PdfGrammar.
    Irony-->>yourprogram: Parsing result.
    
    yourprogram->>PdfSharpDslCore: Define callback for Formula functions.
    yourprogram->>PdfSharpDslCore: Draw()
    PdfSharpDslCore -->>TerraPDF: publish recorded canvas commands.
    PdfSharpDslCore->>yourprogram: call registered formula functions.
    yourprogram-->>PdfSharpDslCore: function result.
    PdfSharpDslCore->>PdfSharpDslCore: execute all instructions from source file.
    PdfSharpDslCore-->>yourprogram: end of draw

Language specification

It's a list of drawing "orders" follow by ';'

All coordinates are specified in 'points'

1 inch = 72.0 points
1 millimeter = 72.0 points / 25.4 ~=> 2.84 points
1 centimeter = 72.0 points / 2.54 ~=> 28.34 points

Formula features

a formula result can be string or double
aritmetic/boolean operations are : +, -, /, *, %, >, <, >=, <=,==, <>, and, or
variable reference : $VarName, and must be declared before with SET VAR VarName=Formula
a formula is a string or a boudle value
Boolean comparison are resolved using double, 0.0 => false, other value => true
string comparison accept only <>, == operators

SET VAR A=2
SET VAR B=3
SET VAR CSquare=$A*$A+$B*$B

Formula can be used in :

each number in [PointLocation, RectLocation, Width, FontSize, startAngle, sweepAngle]
in UserDefineFunction (UDF)
in IF condition

Formula support also customFunction

Register a formula function in your code

var visitor = new PdfDrawerVisitor();
visitor.RegisterFormulaFunction("SUM", (args) => args.Sum(x => Convert.ToDouble(x)));

then use it in formulas

SET VAR CSquare=Sum($A*$A, $B*$B+Sum(1,2,3))

Built-in functions

Available without registration. A function registered with RegisterFormulaFunction under the same name overrides it.

Group	Functions
Math	`Min(a,b,…)`, `Max(a,b,…)`, `Sum(a,b,…)`, `Abs(x)`, `Round(x[,digits])`, `Floor(x)`, `Ceil(x)`, `Sqrt(x)`, `Pow(x,y)`
String	`Upper(s)`, `Lower(s)`, `Len(s)`, `Substr(s,start[,len])`, `Replace(s,a,b)`, `Trim(s)`
Format	`Format(value, fmt)`: .NET format string, invariant culture, e.g. `Format(1284.5,"N2")` → `1,284.50`, `Format(Now(),"yyyy-MM-dd")`
Date	`Now()`, `Today()`: return a `DateTime`. Only usable through `Format`; concatenating one with `+` falls back to `DateTime`'s default, culture-dependent `ToString()`
Logic	`Iif(cond, a, b)`
List	`Count(list)`: number of items of a list

A wrong argument count raises a PdfParserException naming the function.

TextWidth(text) and TextHeight(text[, maxWidth]) measure text in points using the current font (SET FONT), e.g. to size a box to its content:

SET VAR W=TextWidth("Total");
FILLRECT 40,100,$W+20,20;

System variables

Set by the engine, read like any other variable.

Variable	Available	Value
$PAGEWIDTH	everywhere	current page width in points
$PAGEHEIGHT	everywhere	current page height in points
$PAGEINDEX	everywhere	1-based index of the current page
$PAGECOUNT	everywhere (text only, see below)	total number of pages, resolved when the document is published
$ROWINDEX	inside ROWTEMPLATE (free or table)	0-based index of the current iteration
$LASTTEMPLATEHEIGHT	after a ROWTEMPLATE	height in points of the last template, when it did not break the page
$CURSORY	inside FLOW	y where the next element of the flow starts

See also the reserved UDF __ONNEWPAGE (called after each new page). It and the MASTER body can draw with their own font, brush and pen: the ones in use before the page was created are restored afterwards (variables they set stay set).

$PAGECOUNT is only known once every page has been recorded, so it only works inside TITLE/LINETEXT text, e.g. a footer set from __ONNEWPAGE:

TITLE Margin=-18 Text=("page "+$PAGEINDEX+" / "+$PAGECOUNT);

It only supports string concatenation (+); using it in arithmetic or a comparison ($PAGECOUNT-1, $PAGECOUNT>3, …) raises a clear error, since its value does not exist yet while the script runs.

Control flow

# IF ... THEN ... [ELSE IF ... THEN ...]... [ELSE ...] ENDIF
IF $X > 10 THEN
    SET VAR SIZE="big";
ELSE IF $X > 5 THEN
    SET VAR SIZE="medium";
ELSE
    SET VAR SIZE="small";
ENDIF

# FOR var=from TO to [STEP n] DO ... ENDFOR  (bounds are inclusive)
FOR I=0 TO 10 STEP 5 DO ... ENDFOR      # 0, 5, 10
FOR I=10 TO 0 STEP -2 DO ... ENDFOR     # counts down

# WHILE condition DO ... ENDWHILE
SET VAR Y=100;
WHILE $Y < $PAGEHEIGHT-50 DO
    LINE 40,$Y,300,$Y;
    SET VAR Y=$Y+20;
ENDWHILE

ELSE IF (with any whitespace between the words) continues the same IF: the whole chain has a single ENDIF, and conditions are only evaluated until one is true. Breaking change: ELSE followed by a nested IF used to need two ENDIFs; to nest an IF inside an ELSE now, put something (a statement or a # comment) between ELSE and IF.
STEP is an integer formula, 1 by default. A negative step counts down; STEP 0 raises a PdfParserException. A loop whose direction does not reach to (FOR I=5 TO 1) does not run.
WHILE stops with a PdfParserException after 10 000 iterations, so a condition that never becomes false cannot hang the generation.
Loop variables and variables set in a loop stay available after it.

Lists

SET VAR ITEMS=["a","b","c"];
SET VAR FIRST=$ITEMS[0];             # indexes start at 0
SET VAR N=Count($ITEMS);
SET VAR Y=100;
FOREACH ITEM IN $ITEMS DO
    LINETEXT 40,$Y Text=$ITEM;
    SET VAR Y=$Y+16;
ENDFOREACH
SET VAR M=[[1,2],[3,4]];
SET VAR X=$M[1][0];                  # 3

A list literal is [formula, formula, …] ([] is an empty list); items can be any formula, including other lists.
$VAR[index] reads an item; the index is a formula. It can be chained for lists of lists. It only applies to a variable (Names()[0] is not supported; store the result in a variable first). An index that is not a whole number inside the list raises a PdfParserException with the position.
FOREACH var IN formula DO … ENDFOREACH visits the items in order; the formula must give a list.
A formula function registered by the host can return any array or list (not a string): it works with FOREACH, Count and $VAR[index].
A list used in a text ("items: "+$ITEMS) is shown as [a, b, c].

Host data

The host gives the script structured data before drawing:

var visitor = new PdfDrawerVisitor();
visitor.SetData("orders", orders);          // any IEnumerable of objects or dictionaries
visitor.SetData("REPORTTITLE", "Q3 report"); // or a plain value: a preset variable
visitor.Draw(drawer, tree);

LINETEXT 40,40 Text=$REPORTTITLE;
SET VAR Y=80;
FOREACH O IN $orders DO
    LINETEXT 40,$Y Text=($O.customer+": "+Format($O.amount,"N2"));
    SET VAR Y=$Y+16;
ENDFOREACH
ROWTEMPLATE Count=Count($orders) Y=200
    LINETEXT 40,0 Text=$orders[$ROWINDEX].customer;
ENDROWTEMPLATE

$record.field reads a field: from a dictionary by key, from any other object by public property or field (case-insensitive, an exact match wins). It chains with indexes: $orders[0].customer, $order.parent.name, $o.tags[2]; a list item can be a record.
A missing field raises a PdfParserException with the position and, when a name is close, a "Did you mean" (or the list of available fields); a field of an empty value ($o.parent.name with no parent) raises one too.
A dictionary is a record, not a list: FOREACH over one is an error. Any other IEnumerable except a string is a list, so Count, FOREACH and $LIST[i] work on it.
Values keep their .NET type: a DateTime field goes through Format($o.date,"yyyy-MM-dd"), a decimal works in arithmetic.
SetData names are case-sensitive, like variables. The values are copied into the variables at each Draw; the script can overwrite them with SET VAR.

Named styles

STYLE h1
    SET FONT Name="Arial" Size=20 bold;
    SET BRUSH darkblue;
ENDSTYLE
USE h1;
LINETEXT 40,60 Text="Title";

A style is a list of SET PEN, SET BRUSH, SET HBRUSH and SET FONT statements (no SET VAR); USE name; replays them, and they stay set afterwards like any other SET.
Styles are hoisted like UDFs: USE can come before the STYLE definition. USE of an unknown style raises a PdfParserException (with a "Did you mean" suggestion when a style has a similar name); defining the same style twice does too.

INCLUDE

INCLUDE "common.ipdf";

The statements of the file replace the INCLUDE line, so a shared file can define UDFs, STYLEs, MASTERs and variables (SET VAR), and can also draw.
Paths are relative to the file that contains the INCLUDE (so nested includes work from any folder). INCLUDE is only accepted at the top level of a file, not inside a block.
A file is included once per document: a later INCLUDE of the same file is skipped. So every file can INCLUDE the shared files it needs (and still be drawn alone) without defining their UDFs twice.
A missing file, a circular include (a.ipdf -> b.ipdf -> a.ipdf), a nesting deeper than 16 files or a syntax error in the included file raise a PdfParserException; the last one names the file and the line in that file. Run-time errors inside an included file (undefined variable, unknown function, list index…) say at line 2, col 12 of file.ipdf, including from a UDF or MASTER defined there.
Relative IMAGE Source= paths are resolved from the folder of the file that contains the statement.
The visitor resolves the main file's relative paths from its base directory (new PdfDrawerVisitor(baseDirectory, logger), the current directory by default); the console uses the folder of the drawn file. It also works in the source generator. Since an included file is parsed when the document is drawn, the syntax of INCLUDEd files is not checked by the first Parse of the main file.
PdfSharpDslConsole/demo.ipdf is only a list of INCLUDEs of demo-common.ipdf (shared UDFs, styles, master) and demo-includes/NN-feature.ipdf, one file per feature; each of them can be drawn alone, e.g. dotnet run -- demo-includes/11-flow.ipdf.

BARCODE

# BARCODE x,y,w,h Type=code128 Text=formula;
SET BRUSH black;
BARCODE 40,100,260,60 Type=code128 Text="ABC-123";

Draws a Code 128 barcode as vector bars in the current brush. The rectangle includes a quiet zone of 10 modules on each side; scale it so the bars stay wide enough for your scanner.
Code sets B and C are used automatically (runs of 4 or more digits use C, so numeric codes are shorter). The text must be printable ASCII (32 to 126). Other characters, an empty text or a missing/zero width or height raise a PdfParserException.
Works inside ROWTEMPLATE and with DEBUG_RECT. No human-readable text is printed; use LINETEXT under the bars.
QR codes are not available yet (see T25 in tasks.md).

CHART

# CHART bar|line|pie x,y,w,h Data=list [Labels=list] [Colors=list];
CHART bar 40,100,300,200 Data=[12,30,18] Labels=["Q1","Q2","Q3"] Colors=[steelblue];
CHART line 380,100,300,200 Data=[5,-4,8,3] Labels=["a","b","c","d"] Colors=[tomato];
CHART pie 40,340,300,200 Data=[50,30,20] Labels=["Direct","Search","Referral"] Colors=[steelblue,lightgreen,gold];

One chart in one line, drawn with the drawing primitives (FILLRECT, LINE, FILLPIE, LINETEXT) inside the rectangle, so it works in a ROWTEMPLATE, in a FLOW (with absolute coordinates) and with DEBUG_RECT. The current pen, brush and font are restored afterwards; the font family is the current one, at 8 pt.
Data is a list ([12,30,18], a variable, or a host list) or a text "12,30,18". Labels and Colors are lists too (Labels before Colors, both optional).
bar and line have a value axis with round ticks (negative values are drawn below the zero line), one label per item under the axis, and the value above each bar. pie starts at the top and goes clockwise; with Labels a legend with the percentages is drawn at its right.
A color is a name (steelblue, bare in a list or as a text) or "#RRGGBB" / "#AARRGGBB". When there are fewer colors than items they repeat; without Colors a default palette is used (a line chart uses the first color).
Errors raise a PdfParserException: empty data, a value that is not a number, an unknown color (with a suggestion), a pie with a negative value or a zero sum, a rectangle too small for the axes.
A bare color name is now also a valid formula: it is the text of that name (SET VAR C=red;).

Color and Brush

# SET PEN Color Width [Style]
SET PEN black 1;
SET PEN black 1 solid;
SET PEN black 1 dot;
SET PEN black 1 dashdot;
SET PEN black 1 dashdotdot;

**Width** is one of
- number
- Formula


# SET BRUSH Color
SET BRUSH black;

**Color** is one of
- NamedColor : [color list](./README.md#named-color-list)
- HexColor : 0xRGB 
    - sample: 0xFFEEBB

# SET FONT FontName FontSize [FontStyle]
SET FONT Name="Arial" Size=20 bold;

FontName is one of

string
Formula

[FontStyle] is one of

regular (if not specified)
bold
italic
bolditalic
underline
strikeout

Title

Draw text with a specified margin from top or bottom if negative.

# TITLE HAlign=HorizontalAlignment Text="text to draw" 
TITLE HAlign=hcenter Text="TITLE TEST";

# TITLE [MarginTop] HorizontalAlignment "text to draw" 
TITLE Margin=50 HAlign=hcenter Text="My title with margin 50";

HorizontalAlignment : left, hcenter, right [MarginTop] : margin from top if positive, margin from bottom if negative

Ellipse, Rectangle, Line

# ELLIPSE RectLocation
ELLIPSE 5, 5, -5, -5;
# RECT RectLocation
RECT 5, 5, -5, -5;

SET BRUSH orange;
# FILLRECT RectLocation
FILLRECT 250, 100, 50,50;

SET BRUSH green;
# FILLELLIPSE RectLocation
FILLELLIPSE 250, 200, 50,50;

# LINE RectLocation
LINE 100,100, 200, 100;

here Width and Height of RectLocation is X1 and Y1

LINE, ELLIPSE, RECT use PEN (outline)

FILLRECT, FILLELLIPSE use PEN (outline) and BRUSH (fill)

RectLocation is one of

positive number : number of point from left
negative number : number of point from right
formula : supports only "+ - * / ( )"

Pie, FillPie

# PIE RectLocation startAngle sweepAngle
PIE 10,10,120,120 Start=0 Angle=90;
FILLPIE 10,10,120,120 Start=0 Angle=90;

Polygon, FillPolygon

# POLYGON PointLocation PointLocation PointLocation [PointLocation PointLocation PointLocation ...]
POLYGON 300,300, 350,320, 330,350, 240,240;

FILLPOLYGON 100,100, 150,120, 130,150, 240,40;

MoveTo, LineTo

# MOVETO PointLocation
# LINETO PointLocation

MOVETO 300, 200;
LINETO 400, 200;
LINETO 400, 220;
LINETO 300, 200;

New page

# NEWPAGE [PageSize] [PageOrientation];
NEWPAGE ;
NEWPAGE A4 portrait;

[PageSize] is one of

A0, A1, A2, A3, A4, A5, A6, B0, B1, B2, B3, B4, B5, Crown, Demy, DoubleDemy, Elephant, Executive, Folio, Foolscap, GovernmentLetter, LargePost, Ledger, Legal, Letter, Medium, Post, QuadDemy, Quarto, RA0, RA1, RA2, RA3, RA4, RA5, Royal, Size10x14, Statement, STMT, Tabloid, Undefined

[PageSize] is one of

portrait, landscape

MASTER

# MASTER name [MarginTop=formula]
#     statements (usually TITLE, for a header/footer)
# ENDMASTER
MASTER report MarginTop=60
    TITLE Margin=20 Text="ACME report";
    TITLE Margin=-18 Text=("page "+$PAGEINDEX+" / "+$PAGECOUNT);
ENDMASTER

NEWPAGE A4 portrait Master=report;

Masters are hoisted like UDFs, so Master=name can reference one defined later in the file.
The master's statements run, right after __ONNEWPAGE, on the page a NEWPAGE ... Master=name; creates and on any page a ROWTEMPLATE page break creates from it. A later NEWPAGE without Master= has no master, even if the previous page had one — give it Master=name again to keep using it.
MarginTop becomes the default NewPageTopMargin for a ROWTEMPLATE that doesn't specify its own (see ROWTEMPLATE), so content doesn't start under the master's header.
NEWPAGE Master=unknown; raises a PdfParserException.

FLOW

Place content top to bottom without computing y.

# FLOW [Margin=formula] [Top=formula]
#     PARAGRAPH [HAlign=left|right|hcenter] Text=formula;
#     SPACE formula;
#     IMAGE x,y,w,h ...;      # x relative to the left margin, y relative to the cursor
#     TABLE x,y ... ENDTABLE  # same
# ENDFLOW
FLOW Margin=40
    SET FONT Name="Arial" Size=10 regular;
    PARAGRAPH Text="A long text that wraps to the flow width...";
    SPACE 12;
    PARAGRAPH HAlign=right Text=("Now at y = "+$CURSORY);
ENDFLOW

The flow is PageWidth - 2*Margin wide (Margin defaults to 36) and stops Margin above the bottom of the page.
Top is where the first element starts on the first page; it defaults to the active master's MarginTop, or else Margin. Pages created by the flow start there too.
PARAGRAPH wraps its text (current font and brush) and moves the cursor down. A paragraph taller than the space left is split between lines, and continues at the top of the next page. SPACE moves the cursor down (or up if negative) and draws nothing.
IMAGE and TABLE keep their syntax, but x,y are relative to the left margin and the cursor. An image that does not fit moves to the next page; a table breaks between rows (see Top), and the cursor ends below its last row.
A page break made by the flow calls NEWPAGE under the hood, so the current MASTER and __ONNEWPAGE apply to the new page.
$CURSORY is only readable inside a flow. PARAGRAPH/SPACE outside a flow, and a FLOW inside a FLOW, raise a PdfParserException. Other statements can be used in a flow but do not move the cursor.
FLOW works in page points: don't combine it with VIEWSIZE.

Error messages

The console (and PdfDslDiagnostics.FormatParseErrors(parseTree) for hosts) reports errors with line and column, and a suggestion when a name looks like a typo:

Unknown instruction 'LINETXT' at line 12, col 1. Did you mean 'LINETEXT'?
Missing ';' after '1' at line 3, col 14.
Missing 'ENDFOR' for 'FOR' opened at line 5, col 1.
'ENDFOR' at line 9, col 1 does not match 'IF' opened at line 6, col 1. Expected 'ENDIF'.

Run-time errors carry a position too: Variable '$TOTL' is not defined at line 30, col 21. Did you mean '$TOTAL'? and Unknown function 'Uppr' at line 4, col 21. Did you mean 'UPPER'?. Both are PdfParserExceptions (an undefined variable used to be an ArgumentOutOfRangeException).

Image

# IMAGE PointLocation Source=ImageFilePath
IMAGE 100,100 Source="./imageTest.jpg";

# IMAGE PointLocation Data=Base64 encoded image
IMAGE 100,100 Source="data:image/...";

# IMAGE PointLocation,width,height width_height_unit ImageFilePath
IMAGE 320,100,34,34 point Source="./imageTest.jpg";

# IMAGE PointLocation,width,height width_height_unit [cropping] ImageFilePath
IMAGE 100,320,50,50 pixel crop Source="C:\\Samples\\imageTest.jpg";
IMAGE 100,320,50,50 pixel crop Source="C:/Samples/imageTest.jpg";

ImageFilePath : path can be relative or absolute

width_height_unit : 'point' or 'pixel'

when specified image is scale to provided rectangle

[cropping] : don't scale but crop image from provided rectangle

Text

# LINETEXT PointOrRect [hAlign] [vAlign] [Orientation] Text="text"
LINETEXT 42,100 Text="Horizontal text"

LINETEXT 42,100 vertical Text="Horizontal text"
LINETEXT 42,100 left bottom vertical Text="Horizontal text";

[hAlign] is one of

left, right, hcenter

[vAlign] is one of

top, bottom, vcenter

[Orientation] is one of

horizontal, vertical

With a rect location (x,y,w,h), two options handle text that doesn't fit:

LINETEXT 40,100,120,20 Fit=shrink Text="This is a long label";
LINETEXT 40,100,120,20 Overflow=ellipsis Text="This is a long label";

Fit=shrink reduces the font size (down to 4pt) until the text fits the rect.
Overflow=ellipsis truncates the last line that fits the rect's height and appends ….

# TEXT PointOrRect [MaxWidth=formula] Text="text"
TEXT 42,100 Text="Horizontal text"

TEXT 42,100 MaxWidth=50 Text="Horizontal text"

specify MaxWidth to wrap text on multi-lines
this 'TEXT' cannot be align

ROWTEMPLATE

# ROWTEMPLATE Count=formula Y=formula [BorderSize=formula]
# ENDROWTEMPLATE

ROWTEMPLATE Count=3 Y=300 BorderSize=5
    # top line under border
	LINE 0,0,$PAGEWIDTH-20, 0;

	LINE 0,0,50, 50;
    # bottom line under border
	LINE 0,50,$PAGEWIDTH-20, 50;
ENDROWTEMPLATE
# rect include borders
RECT 0,300, $PAGEWIDTH-20, $LASTTEMPLATEHEIGHT

This statement is like a table, but only for row

foreach loop the start point is 0,0
you can access to variable '$ROWINDEX' inside template
after template you can access to '$LASTTEMPLATEHEIGHT' to know the height(in point) of the previous template

[BorderSize]

is to add space between each loop
a space is added on top and at bottom when specified

User Define Function

Udf is like a method in C#, where you can group instructions and reuse multiple times

define function

UDF MyUdf()
LINETEXT 100,100 Text="Horizontal text"
ENDUDF

UDF MyUdf1(X,Y)
LINETEXT $X,$Y Text="Horizontal text"
ENDUDF

udf with parameters

UDF HEXAGONE(X, Y, T)
	SET VAR SQRT3=1.7320;
	LINETEXT $X-$T/2, $Y-$T*$SQRT3/2 HAlign=left VAlign=bottom Text="MOVETO/LINETO";
	MOVETO $X-$T/2, $Y-$T*$SQRT3/2;
	LINETO $X+$T/2, $Y-$T*$SQRT3/2;
	LINETEXT $X+$T/2,$Y-$T*$SQRT3/2 HAlign=left VAlign=bottom Orientation=60 Text="LINETO";
	LINETO $X+$T, $Y+0;
	LINETEXT $X+$T, $Y+0 HAlign=left VAlign=bottom Orientation=120 Text="LINETO";
	LINETO $X+$T/2, $Y+$T*$SQRT3/2;
	LINETEXT $X+$T/2, $Y+$T*$SQRT3/2 HAlign=left VAlign=bottom Orientation=$PI Text="LINETO";

	LINETO $X-$T/2, $Y+$T*$SQRT3/2;
	LINETEXT $X-$T/2, $Y+$T*$SQRT3/2 HAlign=left VAlign=bottom Orientation=240 Text="LINETO";
	LINETO $X-$T, $Y+0;
	LINETEXT $X-$T, $Y+0 HAlign=left VAlign=bottom Orientation=300 Text="LINETO";
	LINETO $X-$T/2, $Y-$T*$SQRT3/2;
ENDUDF

Reserved user function

introduce in 1.0.4

UDF __ONENEWPAGE()
# called on each new page
# add here your custom draw (example draw PageIndex)
    LINETEXT ($PAGEWIDTH/2),$PAGEHEIGHT HAlign=hcenter VAlign=bottom Text=$PAGEINDEX;
ENDUDF

Override UDF in C#

if you redefine 'MyUdf' in C# you can control

var visitor = new PdfDrawerVisitor();
visitor.RegisterCustomUdf("MyUdf", (drawer, argNames, argValues) => {
    //....
    //this udf is called before udf define in '.ipdf' file
    drawer.DrawLineText("Horizontal Text", 100,100,null,null,
        XStringAlignment.Near, XLineAlignment.Far, 
        new TextOrientation(){Orientation=TextOrientationEnum.Horizontal, Angle = 0});
    //return true to override, false to execute also udf define un .ipdf 
    return true;
    });

UDF return values

UDF DOUBLE(X)
    RETURN $X*2;
ENDUDF
SET VAR Y=DOUBLE(21);                 # 42
LINETEXT 40,40 Text=("6! = "+FACT(6));  # a UDF may be recursive

RETURN formula; ends the UDF (even from inside an IF or a loop) and gives the value. It is only allowed in a UDF body; anywhere else raises a PdfParserException.
A formula calls a UDF like any function (NAME(args), the name is not case-sensitive there). Registered functions and built-ins come first: a UDF with the same name is not called from a formula (it still is by CALL).
A UDF used in a formula must RETURN a value, and get exactly its parameters. It may also draw (RECT, LINETEXT, ...): SET VAR NEXTY=BOX(45,$Y,"label"); can draw a box and return the next y. Its variables stay local, as with CALL.
CALL name(...) still works, and a RETURN in the body ends it early (the value is ignored).
A UDF called more than 256 levels deep (endless recursion) raises a PdfParserException.

Call an User Define Function

CALL MyUdf();
# udf with parameters
CALL MyUdf1(100,100);

each parameter can be a Formula

Debugging

DEBUGOPTIONS [GLOBAL|PAGE] Option1 [, Option2];

Scope

GLOBAL (default when omitted) : options apply to the whole document, wherever the statement is written
PAGE : options apply from the statement to the end of the current page, they are reset on each new page (NEWPAGE or ROWTEMPLATE page break)

DEBUGOPTIONS DEBUG_TEXT;
NEWPAGE;
# rule only on this page
DEBUGOPTIONS PAGE DEBUG_RULE;

Available options

DEBUG_TEXT : shows red rect around texts
DEBUG_RECT : shows red rect around each figure (rect, ellipse, pie, polygon)
DEBUG_IMAGE : shows red rect around images
DEBUG_RULE : shows rule on each page
DEBUG_ALL : all of the above
DEBUG_GRID : shows a light grid every 50 points, labelled with its coordinates, to help place elements
DEBUG_ROWTEMPLATE : shows red rect around each iteration and index number of each at topleft rectangle
- text format is "{level}.{index}", where level is > 0 when ROWTEMPLATE is part of another ROWTEMPALTE

Named Color list

aliceblue antiquewhite aqua aquamarine azure beige bisque black blanchedalmond blue blueviolet brown burlywood cadetblue chartreuse chocolate coral cornflowerblue cornsilk crimson cyan darkblue darkcyan darkgoldenrod darkgray darkgreen darkkhaki darkmagenta darkolivegreen darkorange darkorchid darkred darksalmon darkseagreen darkslateblue darkslategray darkturquoise darkviolet deeppink deepskyblue dimgray dodgerblue firebrick floralwhite forestgreen fuchsia gainsboro ghostwhite gold goldenrod gray green greenyellow honeydew hotpink indianred indigo ivory khaki lavender lavenderblush lawngreen lemonchiffon lightblue lightcoral lightcyan lightgoldenrodyellow lightgray lightgreen lightpink lightsalmon lightseagreen lightskyblue lightslategray lightsteelblue lightyellow lime limegreen linen magenta maroon mediumaquamarine mediumblue mediumorchid mediumpurple mediumseagreen mediumslateblue mediumspringgreen mediumturquoise mediumvioletred midnightblue mintcream mistyrose moccasin navajowhite navy oldlace olive olivedrab orange orangered orchid palegoldenrod palegreen paleturquoise palevioletred papayawhip peachpuff peru pink plum powderblue purple red rosybrown royalblue saddlebrown salmon sandybrown seagreen seashell sienna silver skyblue slateblue slategray snow springgreen steelblue tan teal thistle tomato transparent turquoise violet wheat white whitesmoke yellow yellowgreen

Dependencies :

this package is build on top of

PDF :
- TerraPDF : https://www.nuget.org/packages/TerraPDF
Parsers :
- Irony : https://github.com/IronyProject/Irony

Source Generator

You can generate C# from PDF DSL, in order to have "hard coded" PDF. You can generate C# code of UDF (user define function)

is not yet available

Name		Name	Last commit message	Last commit date
.github/workflows		.github/workflows
PdfSharpDsl.Language		PdfSharpDsl.Language
PdfSharpDslConsole		PdfSharpDslConsole
PdfSharpDslCore.Generator		PdfSharpDslCore.Generator
PdfSharpDslCore		PdfSharpDslCore
_build		_build
pdfsharpdslTests		pdfsharpdslTests
scripts		scripts
.gitattributes		.gitattributes
.gitignore		.gitignore
.runsettings		.runsettings
CHANGELOG.md		CHANGELOG.md
Directory.Build.props		Directory.Build.props
Directory.Build.targets		Directory.Build.targets
Directory.Packages.props		Directory.Packages.props
LICENSE.md		LICENSE.md
README.md		README.md
bnf_and_pdf.sln		bnf_and_pdf.sln
global.json		global.json
tasks.md		tasks.md

Latest commit

History

Folders and files

Repository files navigation