added functionality for receiving and parsing stext from page

Cargo.toml

@@ -47,6 +47,8 @@ mupdf-sys = { version = "0.4.2", path = "mupdf-sys" }
 once_cell = "1.3.1"
 num_enum = "0.7.0"
 bitflags = "2.0.2"
+serde = { version = "1.0.201", features = ["derive"] }
+serde_json = "1.0.117"
 
 [dependencies.font-kit]
 version = "0.12.0"

examples/extract_stext.rs

@@ -0,0 +1,33 @@
+use std::io;
+
+fn main() {
+ // cargo run --example extract_stext
+ let mut path_to_doc = String::new();
+ println!("Enter a path to document: ");
+ io::stdin()
+ .read_line(&mut path_to_doc)
+ .expect("Failed to read line");
+ let doc = mupdf::document::Document::open(path_to_doc.trim()).unwrap();
+ let page = doc.load_page(0).unwrap();
+ match page.stext_page_as_json_from_page(1.0) {
+ Ok(stext_json) => {
+ let stext_page: serde_json::Result<mupdf::page::StextPage> =
+ serde_json::from_str(stext_json.as_str());
+ match stext_page {
+ Ok(res) => {
+ for block in res.blocks {
+ if block.r#type.eq("text") {
+ for line in block.lines {
+ println!("{:?}", &line.text);
+ }
+ }
+ }
+ }
+ Err(err) => {
+ println!("stext_page parsing error: {:?}", &err);
+ }
+ }
+ }
+ Err(_err) => {}
+ }
+}

mupdf-sys/wrapper.c

@@ -889,6 +889,34 @@ fz_buffer *mupdf_page_to_html(fz_context *ctx, fz_page *page, mupdf_error_t **er
  return buf;
 }
 
+fz_buffer *mupdf_stext_page_as_json_from_page(fz_context *ctx, fz_page *page, float scale, mupdf_error_t **errptr)
+{
+ fz_buffer *buf = NULL;
+ fz_output *out = NULL;
+ fz_stext_page *stext_page = NULL;
+ fz_var(stext_page);
+ fz_var(buf);
+ fz_var(out);
+ fz_try(ctx)
+ {
+ stext_page = fz_new_stext_page_from_page(ctx, page, NULL);
+ buf = fz_new_buffer(ctx, 8192);
+ out = fz_new_output_with_buffer(ctx, buf);
+ fz_print_stext_page_as_json(ctx, out, stext_page, scale);
+ fz_close_output(ctx, out);
+ }
+ fz_always(ctx)
+ {
+ fz_drop_output(ctx, out);
+ fz_drop_stext_page(ctx, stext_page);
+ }
+ fz_catch(ctx)
+ {
+ mupdf_save_error(ctx, errptr);
+ }
+ return buf;
+}
+
 fz_buffer *mupdf_page_to_xhtml(fz_context *ctx, fz_page *page, mupdf_error_t **errptr)
 {
  fz_buffer *buf = NULL;

src/page.rs

@@ -3,6 +3,8 @@ use std::io::Read;
 use std::ptr;
 use std::slice;
 
+use serde::{Deserialize, Serialize};
+
 use mupdf_sys::*;
 
 use crate::{
@@ -235,6 +237,20 @@ impl Page {
  Ok(out)
  }
 
+ pub fn stext_page_as_json_from_page(&self, scale: f32) -> Result<String, Error> {
+ let mut buf = unsafe {
+ let inner = ffi_try!(mupdf_stext_page_as_json_from_page(
+ context(),
+ self.inner,
+ scale
+ ));
+ Buffer::from_raw(inner)
+ };
+ let mut res = String::new();
+ buf.read_to_string(&mut res).unwrap();
+ Ok(res)
+ }
+
  pub fn to_xhtml(&self) -> Result<String, Error> {
  let mut buf = unsafe {
  let inner = ffi_try!(mupdf_page_to_xhtml(context(), self.inner));
@@ -359,10 +375,83 @@ impl Iterator for LinkIter {
  }
 }
 
+#[derive(Deserialize, Serialize, Debug, Clone)]
+pub struct Font {
+ pub name: String,
+ pub family: String,
+ pub weight: String,
+ pub style: String,
+ pub size: u32,
+}
+
+#[derive(Deserialize, Serialize, Debug, Clone)]
+pub struct BBox {
+ pub x: u32,
+ pub y: u32,
+ pub w: u32,
+ pub h: u32,
+}
+
+#[derive(Deserialize, Serialize, Debug, Clone)]
+pub struct Line {
+ pub wmode: u32,
+ pub bbox: BBox,
+ pub font: Font,
+ pub x: u32,
+ pub y: u32,
+ pub text: String,
+}
+
+#[derive(Deserialize, Serialize, Debug, Clone)]
+pub struct Block {
+ pub r#type: String,
+ pub bbox: BBox,
+ pub lines: Vec<Line>,
+}
+
+// StructuredText
+#[derive(Deserialize, Serialize, Debug, Clone)]
+pub struct StextPage {
+ pub blocks: Vec<Block>,
+}
+
 #[cfg(test)]
 mod test {
+ use crate::page::StextPage;
  use crate::{Document, Matrix};
 
+ #[test]
+ fn test_get_stext_page_as_json() {
+ let path_to_doc = std::env::current_dir()
+ .unwrap()
+ .join("tests")
+ .join("files")
+ .join("dummy.pdf");
+ let doc = Document::open(path_to_doc.to_str().unwrap()).unwrap();
+ let page = doc.load_page(0).unwrap();
+ match page.stext_page_as_json_from_page(1.0) {
+ Ok(stext_json) => {
+ let stext_page: serde_json::Result<StextPage> =
+ serde_json::from_str(stext_json.as_str());
+ match stext_page {
+ Ok(res) => {
+ for block in res.blocks {
+ if block.r#type.eq("text") {
+ for line in block.lines {
+ assert_eq!(&line.text, &"Dummy PDF file".to_string());
+ }
+ }
+ }
+ }
+ Err(err) => {
+ println!("stext_page parsing error: {:?}", &err);
+ }
+ }
+ }
+ Err(_err) => {}
+ }
+ }
+
  #[test]
  fn test_page_to_svg() {
  let doc = Document::open("tests/files/dummy.pdf").unwrap();
@@ -454,20 +543,20 @@ mod test {
  [Quad {
  ul: Point {
  x: 56.8,
- y: 69.32512
+ y: 69.32512,
  },
  ur: Point {
  x: 115.85405,
- y: 69.32512
+ y: 69.32512,
  },
  ll: Point {
  x: 56.8,
- y: 87.311844
+ y: 87.311844,
  },
  lr: Point {
  x: 115.85405,
- y: 87.311844
- }
+ y: 87.311844,
+ },
  }]
  );
 

src/text_page.rs

@@ -115,7 +115,7 @@ impl TextBlock<'_> {
 
  pub fn lines(&self) -> TextLineIter {
  unsafe {
- if self.inner.type_ == FZ_STEXT_BLOCK_TEXT as _ {
+ if self.inner.type_ == FZ_STEXT_BLOCK_TEXT as i32 {
  return TextLineIter {
  next: self.inner.u.t.first_line,
  _marker: PhantomData,
@@ -130,7 +130,7 @@ impl TextBlock<'_> {
 
  pub fn ctm(&self) -> Option<Matrix> {
  unsafe {
- if self.inner.type_ == FZ_STEXT_BLOCK_IMAGE as _ {
+ if self.inner.type_ == FZ_STEXT_BLOCK_IMAGE as i32 {
  return Some(self.inner.u.i.transform.into());
  }
  }
@@ -139,7 +139,7 @@ impl TextBlock<'_> {
 
  pub fn image(&self) -> Option<Image> {
  unsafe {
- if self.inner.type_ == FZ_STEXT_BLOCK_IMAGE as _ {
+ if self.inner.type_ == FZ_STEXT_BLOCK_IMAGE as i32 {
  let inner = self.inner.u.i.image;
  fz_keep_image(context(), inner);
  return Some(Image::from_raw(inner));