seehuhn / go-pdf Goto Github PK

It would be really useful to be able to embed an SVG in a PDF.

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/5

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/4

I can merge multiple pdf files into one pdf file?

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/37

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/36

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/16

I tried this

func extractText(fname string) error {
    fd, err := os.Open(fname)
    if err != nil {
        return err
    }
    defer fd.Close()

    r, err := pdf.NewReader(fd, nil)
    if err != nil {
        return err
    }

    contents := reader.New(r, nil)
    contents.Text = func(text string) error {
        fmt.Print(text)
        return nil
    }

    pages := pagetree.NewIterator(r)
    pageNo := 0
    pages.All()(func(_ pdf.Reference, pageDict pdf.Dict) bool {
        fmt.Println("Page", pageNo)

        err := contents.ParsePage(pageDict, matrix.Identity)
        if err != nil {
            log.Fatal(err)
        }

        pageNo++
        return true
    })
    return nil
}

and was hoping this would extract text for machine generated PDFs or PDF with OCR information added - but it prints nothing on all the PDFs I tried. What am I missing?

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/2

How to import an existing pdf file?
How to automatically wrap long text?

I tried to open an encrypted PDF created with the latest version of Adobe Acrobat and I got the error: encryption dictionary: standard security handler: invalid Encrypt.O.

Test file (password is password):
password_protected.pdf

I tested with both v0.3.4 and v0.3.5-0.20230822001153-4ee04e1da286 / 4ee04e1.

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/15

Something seems problematic reading the file ending:

Test POA v2.pdf

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/35

I used (*document.MultiPage).AddPage() and (*document.Page).SetPageSize to create multiple pages in a document.MultiPage with different sizes. If text is added to these pages with (*graphics.Page).TextStart, (*graphics.Page).TextSetFont, (*graphics.Page).TextFirstLine, (*graphics.Page).TextShow, and (*graphics.Page).TextEnd, the text is stretched on each page based on each page's aspect ratio.

func TraverseTree(tree *outline.Tree, level int) {
if tree == nil {
return
}
var num uint32
var ref gopdf.Reference
n := ""
if a, ok := tree.Action["D"]; ok {
arr := a.(gopdf.Array)
ref = arr[0].(gopdf.Reference)
num = ref.Number()
n = strconv.FormatInt(int64(num), 10)
}
fmt.Printf("%s%s:%d:%s:%s\n", strings.Repeat("\t", level), tree.Title, num, n, ref.String()) // Print the Title, or do whatever you want to do with Open/Action
for _, child := range tree.Children {
TraverseTree(child, level+1)
}
}

I just can’t figure out how to draw an image in the place I need, could you please give me an example.
The example shown draws it at the bottom left.

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/32

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/12

Is there a way to replace text using this library?

Great library, I fall back on ghostscript on pdf corruption

I get a string "fish" panic when invoking content.ForAllText
seehuhn.de/go/pdf/content.ForAllText.func2.1
/Users/foxyboy/go/pkg/mod/seehuhn.de/go/[email protected]/content/extract.go:223

if I use my own code I change line 223 to:

panic(perrors.ErrorfPF("unknown graphics state key: %q", key))
// import "github.com/haraldrudell/parl/perrors"

In a 2006 pdf, “SM” is used
Attached is the erroring pdf

some thoughts:
it is good software engineering practice to only use error values as panic arguments
ErrorfPF contains a stack trace that can be printed using perrors.Long
perrors.Short prints message with a short code reference
error messages should be actionable

it is better to return error rather than panic for anything that is recoverable, such as corrupt file data

The object model is peculiar: generic pdf.Object values are handed to top-level functions
of sub-packages. What value goes with which function is unclear. It is therefore difficult to understand how the library is to be used.
Objects should be based on world concepts, so Document is the top-level
Document then has methods to return number of pages, pages and so forth. In this way, library usage would be obvious.
What in 0.3.6 is different packages should be different types (struct/methods) of the same package. It is important that api is real-world object and usage, even if an internal disposition comes from the pdf specification

strings like “MediaBox” should be const that can be searched for, used by api consumers and aren’t misspelled. I also ran into MediaBox value being Array or Rectangle. With supporting multiple versions, shims is a good approach because it tags code with why it exists

pagetree should only be a separate package if its intended to be used separate from Open. It is better a type of the same package

Note that in Go, methods for the same object can go in different files, so many methods is not a problem

New-functions enhance encapsulation. A New function should not store pointers to its created object even in the object itself, or launch threads or return anything other than a single pointer to the value created

If a function literal ends up being used, in Go that means another struct needs to be created with the literal function as a method

aspect.pdf

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/33

Tracking issue for:

• https://github.com/seehuhn/go-pdf/security/code-scanning/10