Add Social_Bookmark_extraction.R for parsing social bookmark HTML exports

Social_Bookmark_extraction.R

@@ -0,0 +1,69 @@
+# Social Bookmark Extraction Script
+# Extracts links and nearby title/snippet text from a local HTML export.
+
+library(rvest)
+library(dplyr)
+library(stringr)
+
+# ---- Input ----
+# Update this path to point to your exported HTML file.
+html_path <- "path/to/your/bookmarks_export.html"
+
+# ---- Optional platform-specific pattern ----
+# Examples:
+# twitter_pattern  <- "status"
+# linkedin_pattern <- "feed/update"
+# facebook_pattern <- "posts|story.php|permalink"
+post_pattern <- "status|feed/update|posts|story.php|permalink"
+
+# ---- Read HTML ----
+page <- read_html(html_path)
+
+# Use file location as base URL for resolving any relative links.
+base_url <- paste0("file://", normalizePath(html_path, winslash = "/", mustWork = FALSE))
+
+# ---- Extract all <a> tags, not class-dependent ----
+all_links <- page %>% html_elements("a")
+
+extract_nearest_text <- function(node) {
+  # Try parent span/div first
+  parent_text <- node %>%
+    html_element(xpath = "./ancestor::*[self::span or self::div][1]") %>%
+    html_text2(trim = TRUE)
+
+  # Fallback to nearest meaningful ancestor containing visible text
+  if (is.na(parent_text) || parent_text == "") {
+    parent_text <- node %>%
+      html_element(xpath = "./ancestor::*[normalize-space(text()) != ''][1]") %>%
+      html_text2(trim = TRUE)
+  }
+
+  if (is.na(parent_text) || str_squish(parent_text) == "") {
+    return("No Title Found")
+  }
+
+  str_squish(parent_text)
+}
+
+bookmarks_cleaned <- tibble(
+  url_raw = html_attr(all_links, "href"),
+  title = vapply(all_links, extract_nearest_text, character(1))
+) %>%
+  filter(!is.na(url_raw), url_raw != "") %>%
+  mutate(
+    # Convert relative URLs to absolute
+    url = url_absolute(url_raw, base = base_url),
+    # Remove tracking/query params (everything after '?')
+    url = str_remove(url, "\\?.*$")
+  ) %>%
+  # Keep likely social post links only
+  filter(str_detect(url, regex(post_pattern, ignore_case = TRUE))) %>%
+  mutate(
+    title = if_else(is.na(title) | str_squish(title) == "", "No Title Found", title)
+  ) %>%
+  distinct(url, .keep_all = TRUE) %>%
+  select(title, url)
+
+print(bookmarks_cleaned)
+
+write.csv(bookmarks_cleaned, "bookmarks_cleaned.csv", row.names = FALSE)