From 6c31131cdd8caa62cc06db42935009bd82cf81a3 Mon Sep 17 00:00:00 2001
From: philipp <philipp@hofer.link>
Date: Thu, 15 Feb 2024 15:07:36 +0100
Subject: [PATCH] enable function to move paragraphs into absaetze

---
 Cargo.lock             |  1 +
 Cargo.toml             |  1 +
 src/config.rs          | 14 ++++++++++++++
 src/law/responsible.rs |  1 +
 src/paragraph/mod.rs   | 44 ++++++++++++++++++++++++++++++++++++++++++
 5 files changed, 61 insertions(+)

diff --git a/Cargo.lock b/Cargo.lock
index fcf1bf2..1deae59 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -583,6 +583,7 @@ dependencies = [
  "env_logger",
  "log",
  "pretty_assertions",
+ "regex",
  "roxmltree",
  "serde",
  "serde_json",
diff --git a/Cargo.toml b/Cargo.toml
index fca36a0..3040264 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -17,6 +17,7 @@ tqdm = "0.6"
 toml = "0.8"
 clap = { version = "4.5.0", features = ["derive"] }
 directories = "5.0"
+regex = "1.10"
 
 [dev-dependencies]
 pretty_assertions = "1.4"
diff --git a/src/config.rs b/src/config.rs
index d6da168..a508b8c 100644
--- a/src/config.rs
+++ b/src/config.rs
@@ -82,6 +82,10 @@ impl Config {
             parser.add_string_to_replace(&to_replace.find, &to_replace.replace_with);
         }
 
+        if config.parser.move_para_headers_into_content {
+            parser.move_para_headers_into_content();
+        }
+
         Ok((config.law.id, builder, parser))
     }
 }
@@ -101,6 +105,16 @@ struct Classifier {
 
 #[derive(Debug, Deserialize, Default)]
 struct ParserConfig {
+    /// e.g. used in EheG to transform `<ueberschrift typ="para" ct="text" halign="c">§
+    /// 6</ueberschrift>` into
+    /// `
+    /// <absatz typ="abs" ct="text" halign="j">
+    /// <gldsym>§ 1.</gldsym>
+    /// text...
+    /// </absatz>
+    ///`
+    #[serde(default)] //okay to not have this part in the config
+    move_para_headers_into_content: bool,
     #[serde(default)] //okay to not have this part in the config
     remove_strings: Vec<String>,
     #[serde(default)] //okay to not have this part in the config
diff --git a/src/law/responsible.rs b/src/law/responsible.rs
index 13f28ef..e0fa480 100644
--- a/src/law/responsible.rs
+++ b/src/law/responsible.rs
@@ -63,6 +63,7 @@ pub fn starts_with_letter(_classifier_name: &str, instance_name: &str) -> bool {
 
 pub fn starts_with_uppercaseletter(_classifier_name: &str, instance_name: &str) -> bool {
     instance_name.starts_with(|c: char| c.is_ascii_uppercase())
+        && instance_name.chars().nth(0) != Some('I')
         && (instance_name.chars().nth(1) == Some('.') || instance_name.chars().nth(1) == Some(')'))
 }
 
diff --git a/src/paragraph/mod.rs b/src/paragraph/mod.rs
index f1e031c..0b73f07 100644
--- a/src/paragraph/mod.rs
+++ b/src/paragraph/mod.rs
@@ -17,6 +17,7 @@
 //! Deals with getting all paragraphs for a given law text
 mod parser;
 
+use regex::Regex;
 use std::{
     fs,
     hash::{DefaultHasher, Hash, Hasher},
@@ -35,6 +36,7 @@ use self::parser::Risdok;
 pub struct Parser {
     remove: Vec<String>,
     replace: Vec<(String, String)>,
+    move_para_headers_into_content: bool,
 }
 
 impl Default for Parser {
@@ -48,9 +50,14 @@ impl Parser {
         Self {
             remove: Vec::new(),
             replace: Vec::new(),
+            move_para_headers_into_content: false,
         }
     }
 
+    pub fn move_para_headers_into_content(&mut self) {
+        self.move_para_headers_into_content = true;
+    }
+
     pub fn add_string_to_remove(&mut self, data: &str) {
         self.remove.push(data.into());
     }
@@ -78,8 +85,45 @@ impl Parser {
             xml = xml.replace(search, replace);
         }
 
+        let xml = if self.move_para_headers_into_content {
+            Self::do_move_para_headers_into_content(xml)
+        } else {
+            xml
+        };
+
         Risdok::from_str(&xml, builder)
     }
+
+    fn do_move_para_headers_into_content(xml: String) -> String {
+        let mut result = String::from(&xml);
+        let ueberschrift_regex = Regex::new(
+            "<ueberschrift typ=\"[^\"]*\" ct=\"[^\"]*\" halign=\"[^\"]*\">(§.*?)</ueberschrift>",
+        )
+        .unwrap();
+        let absatz_regex =
+            Regex::new("<absatz typ=\"[^\"]*\" ct=\"[^\"]*\" halign=\"[^\"]*\">").unwrap();
+
+        // Find all matches for <ueberschrift> tags and iterate over them in reverse to avoid messing up the indices
+        for cap in ueberschrift_regex.captures_iter(&xml) {
+            let ueberschrift_content = &cap[1];
+
+            // Check if there's an <absatz> following the <ueberschrift>
+            if let Some(absatz_match) = absatz_regex.find(&result[cap.get(0).unwrap().end()..]) {
+                // Calculate the insertion point for the <gldsym> tag
+                let insert_point =
+                    cap.get(0).unwrap().end() + absatz_match.start() + absatz_match.as_str().len();
+                // Insert the <gldsym> tag with the ueberschrift content into the result string
+                result.insert_str(
+                    insert_point,
+                    &format!("<gldsym>{}</gldsym>", ueberschrift_content),
+                );
+            }
+
+            // Remove the <ueberschrift> tag from the result string
+            result.replace_range(cap.get(0).unwrap().range(), "");
+        }
+        result
+    }
 }
 
 fn fetch(url: &str) -> Result<String, Error> {