From cd747ac41b937e5cafca5146cf1fb55514ec09f0 Mon Sep 17 00:00:00 2001
From: Wouter Coppieters <wouter.coppieters@employmenthero.com>
Date: Sun, 18 Feb 2024 14:04:56 +1300
Subject: [PATCH] v0.1.10, support empty tableCells, and tableCells containing
 just hard breaks'

---
 Cargo.lock             |   2 +-
 Cargo.toml             |   2 +-
 README.md              |  16 +--
 docs/index.html        |   2 +-
 src/adf_builder.rs     |   2 +-
 src/adf_structure.rs   |   2 +-
 src/extractor.rs       |  22 ++++
 src/tests/mod.rs       |   1 +
 src/tests/tables.rs    | 247 +++++++++++++++++++++++++++++++++++++++++
 src/types/node_list.rs |   3 +-
 10 files changed, 285 insertions(+), 14 deletions(-)
 create mode 100644 src/tests/tables.rs

diff --git a/Cargo.lock b/Cargo.lock
index 6414fbb..6783a5a 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -260,7 +260,7 @@ dependencies = [
 
 [[package]]
 name = "htmltoadf"
-version = "0.1.9"
+version = "0.1.10"
 dependencies = [
  "clap",
  "ego-tree",
diff --git a/Cargo.toml b/Cargo.toml
index adae989..6a6d0ad 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "htmltoadf"
-version = "0.1.9"
+version = "0.1.10"
 edition = "2021"
 license = "MIT"
 description = "An HTML to Atlassian Document Format (ADF) converter"
diff --git a/README.md b/README.md
index 510b020..f432545 100644
--- a/README.md
+++ b/README.md
@@ -38,7 +38,7 @@ $ html2adf -h
 ```
 
 ```
-htmltoadf 0.1.9
+htmltoadf 0.1.10
 An HTML to Atlassian Document Format (ADF) converter
 
 USAGE:
@@ -56,20 +56,20 @@ OPTIONS:
 ### Install Binary from Crates.io with `cargo install`
 ```
 $ cargo install htmltoadf
-    installing htmltoadf v0.1.9 (/usr/src/html2adf)
+    installing htmltoadf v0.1.10 (/usr/src/html2adf)
     Updating crates.io index
  Downloading crates ...
   Downloaded lock_api v0.4.6
 --snip--
-      Compiling htmltoadf v0.1.9
+      Compiling htmltoadf v0.1.10
     Finished release [optimized] target(s) in 1m 42s
   Installing ~/.cargo/bin/htmltoadf
-   Installed package `htmltoadf v0.1.9` (executable `html2adf`)
+   Installed package `htmltoadf v0.1.10` (executable `html2adf`)
 ```
 
 ### Download Binary file from Github
 Pre-built binaries can be downloaded from here:
-https://github.com/wouterken/htmltoadf/releases/tag/0.1.9
+https://github.com/wouterken/htmltoadf/releases/tag/0.1.10
 
 ### Docker Image
 **Docker Repo:**
@@ -79,10 +79,10 @@ https://hub.docker.com/r/wouterken/html2adf
 **Usage**
 
 ```bash
-$ echo "<h1>Hello world<p>Test</p></h1>" | docker run --rm -i wouterken/html2adf:0.1.9
+$ echo "<h1>Hello world<p>Test</p></h1>" | docker run --rm -i wouterken/html2adf:0.1.10
 {"version":1,"type":"doc","content":[{"type":"heading","attrs":{"level":1},"content":[{"type":"text","text":"Hello world"},{"type":"text","text":"Test"}]}]}
 
-$ echo "<h1>Hello world<p>Test</p></h1>" | docker run --rm -i wouterken/html2adf:0.1.9 | jq
+$ echo "<h1>Hello world<p>Test</p></h1>" | docker run --rm -i wouterken/html2adf:0.1.10 | jq
 {
   "version": 1,
   "type": "doc",
@@ -115,7 +115,7 @@ $ echo "<h1>Hello world<p>Test</p></h1>" | docker run --rm -i wouterken/html2adf
 
 ```toml
 [dependencies]
-htmltoadf = "0.1.9"
+htmltoadf = "0.1.10"
 ```
 
 **Code**
diff --git a/docs/index.html b/docs/index.html
index 9310411..a7d54f2 100644
--- a/docs/index.html
+++ b/docs/index.html
@@ -79,7 +79,7 @@
 
 </style>
 <script defer type="module">
- import init, {convert} from "https://unpkg.com/htmltoadf@0.1.9/htmltoadf.js";
+ import init, {convert} from "https://unpkg.com/htmltoadf@0.1.10/htmltoadf.js";
 
  let editor;
 
diff --git a/src/adf_builder.rs b/src/adf_builder.rs
index d9e93b7..d014b10 100644
--- a/src/adf_builder.rs
+++ b/src/adf_builder.rs
@@ -11,7 +11,7 @@ use scraper::ElementRef;
 use scraper::Html;
 use serde_json::{Map, Value};
 
-static VALID_EMPTY_TYPES: [&str; 4] = ["hr", "iframe", "img", "br"];
+static VALID_EMPTY_TYPES: [&str; 5] = ["hr", "iframe", "img", "br", "td"];
 
 /**
 * The main procedure for our ADF Builder.
diff --git a/src/adf_structure.rs b/src/adf_structure.rs
index dc28e4f..dc7bcba 100644
--- a/src/adf_structure.rs
+++ b/src/adf_structure.rs
@@ -58,7 +58,7 @@ lazy_static! {
     ),
     (
       String::from("tableCell"),
-      AdfPermittedChildren::any(&["codeBlock", "blockCard", "paragraph", "bulletList", "mediaSingle", "orderedList", "heading", "panel", "blockquote", "rule", "mediaGroup", "decisionList", "taskList", "extension", "embedCard", "nestedExpand"])
+      AdfPermittedChildren::any(&["codeBlock", "blockCard", "paragraph", "bulletList", "mediaSingle", "orderedList", "heading", "panel", "blockquote", "rule", "mediaGroup", "decisionList", "taskList", "extension", "embedCard", "nestedExpand", "hardBreak"])
     ),
     (
       String::from("doc"),
diff --git a/src/extractor.rs b/src/extractor.rs
index 75d4196..eb97c05 100644
--- a/src/extractor.rs
+++ b/src/extractor.rs
@@ -1,4 +1,5 @@
 use ego_tree::iter::Edge;
+use ego_tree::NodeRef;
 use regex::Regex;
 use scraper::Node;
 use scraper::{ElementRef, Html};
@@ -32,6 +33,18 @@ pub fn squish_surrounding_whitespace(input: &str) -> String {
     re.replace_all(input, " ").to_string()
 }
 
+pub fn has_text_node(node: NodeRef<Node>) -> bool {
+    node.children().any(|node| {
+        if let Some(element) = node.value().as_element() {
+            element.name() == "br" || has_text_node(node)
+        }
+        else if let Some(text_node) = node.value().as_text() {
+            !text_node.text.trim().is_empty()
+        } else {
+            false
+        }
+    })
+}
 /**
  * We parse a raw scraper::HTML and return a
  * list of leaf doc nodes  (each with a linked list pointer to the root)
@@ -63,6 +76,15 @@ pub fn extract_leaves(fragment: &Html) -> Vec<DocNode> {
                             text: "".trim().to_owned(),
                             node,
                         })
+                    } else if element.value().name() == "td" {
+                        let has_text_node = has_text_node(node);
+                        if !has_text_node {
+                            leaf_nodes.push(DocNode {
+                                name: "td",
+                                text: "".trim().to_owned(),
+                                node,
+                            })
+                        }
                     }
                 } else if let Node::Text(text_node) = node.value() {
                     if let Some(parent) = node.parent().and_then(ElementRef::wrap) {
diff --git a/src/tests/mod.rs b/src/tests/mod.rs
index 2df0d59..0b4ca67 100644
--- a/src/tests/mod.rs
+++ b/src/tests/mod.rs
@@ -7,6 +7,7 @@ mod image;
 mod lists;
 mod marks;
 mod paragraphs;
+mod tables;
 use crate::convert_html_str_to_adf_str;
 
 #[allow(dead_code)]
diff --git a/src/tests/tables.rs b/src/tests/tables.rs
new file mode 100644
index 0000000..eea4636
--- /dev/null
+++ b/src/tests/tables.rs
@@ -0,0 +1,247 @@
+#[allow(unused_imports)]
+use super::assert_output_json_eq;
+
+#[allow(unused_imports)]
+use serde_json::json;
+
+#[cfg(test)]
+#[test]
+fn test_empty_cell() {
+    assert_output_json_eq(
+        r"<div><table ><tbody>
+            <tr><td >A</td><td >B</td><td >C</td></tr>
+            <tr><td >value 1</td><td ></td><td >value 2</td></tr>
+            </tbody></table>
+        </div>",
+        json!({
+          "version": 1,
+          "type": "doc",
+          "content": [
+            {
+              "type": "table",
+              "content": [
+                {
+                  "type": "tableRow",
+                  "content": [
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "A"
+                            }
+                          ]
+                        }
+                      ]
+                    },
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "B"
+                            }
+                          ]
+                        }
+                      ]
+                    },
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "C"
+                            }
+                          ]
+                        }
+                      ]
+                    }
+                  ]
+                },
+                {
+                  "type": "tableRow",
+                  "content": [
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "value 1"
+                            }
+                          ]
+                        }
+                      ]
+                    },
+                    {
+                      "type": "tableCell"
+                    },
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "value 2"
+                            }
+                          ]
+                        }
+                      ]
+                    }
+                  ]
+                }
+              ]
+            },
+            {
+              "type": "paragraph",
+              "content": [
+                {
+                  "type": "text",
+                  "text": " "
+                }
+              ]
+            }
+          ]
+        }),
+    );
+}
+
+#[test]
+/**
+ * Nested Paragraphs are flattened
+ */
+ fn test_hard_break_in_cell() {
+    assert_output_json_eq(
+        r"<div><table ><tbody>
+            <tr><td >A</td><td >B</td><td >C</td></tr>
+            <tr><td >value 1</td><td ><br/></td><td >value 2</td></tr>
+            </tbody></table>
+        </div>",
+        json!({
+          "version": 1,
+          "type": "doc",
+          "content": [
+            {
+              "type": "table",
+              "content": [
+                {
+                  "type": "tableRow",
+                  "content": [
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "A"
+                            }
+                          ]
+                        }
+                      ]
+                    },
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "B"
+                            }
+                          ]
+                        }
+                      ]
+                    },
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "C"
+                            }
+                          ]
+                        }
+                      ]
+                    }
+                  ]
+                },
+                {
+                  "type": "tableRow",
+                  "content": [
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "value 1"
+                            }
+                          ]
+                        }
+                      ]
+                    },
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "hardBreak"
+                            }
+                          ]
+                        }
+                      ]
+                    },
+                    {
+                      "type": "tableCell",
+                      "content": [
+                        {
+                          "type": "paragraph",
+                          "content": [
+                            {
+                              "type": "text",
+                              "text": "value 2"
+                            }
+                          ]
+                        }
+                      ]
+                    }
+                  ]
+                }
+              ]
+            },
+            {
+              "type": "paragraph",
+              "content": [
+                {
+                  "type": "text",
+                  "text": " "
+                }
+              ]
+            }
+          ]
+        }),
+    );
+}
diff --git a/src/types/node_list.rs b/src/types/node_list.rs
index 068f0ad..c63614c 100644
--- a/src/types/node_list.rs
+++ b/src/types/node_list.rs
@@ -167,7 +167,8 @@ impl NodeList {
                     String::from("media"),
                     String::from("emoji"),
                     String::from("rule"),
-                    String::from("hardBreak")
+                    String::from("hardBreak"),
+                    String::from("tableCell")
                 ];
 
                 if !node.children.is_empty() {