R读取MS Word文档时latin1编码错误及gsub代码问题求助

Hey there! Let's work through your Latin1 encoding error and gsub double-quote issue step by step—here's a practical breakdown tailored to your workflow:

1. Tackle the Latin1 Encoding Core Issue #

First off, encoding errors usually pop up because your Word docs contain characters Latin1 can't represent, or you're reading the text with the wrong encoding specified. Let's fix this at the source when reading your docs:

I recommend using the officer package (it's more reliable for modern Word files than older tools). Here's how to read and convert encoding properly:

library(officer)
library(dplyr)

# Read a single Word doc
doc <- read_docx("your_file.docx")
raw_text <- docx_summary(doc)$content

# Convert Latin1 to UTF-8 (drop unconvertible chars, or use "?" instead of "")
cleaned_encoding <- iconv(raw_text, from = "latin1", to = "UTF-8", sub = "")

The sub = "" argument removes characters that can't be converted—swap it for "?" if you'd rather mark them instead of deleting.

2. Fix the Gsub Double-Quote Problem #

The consecutive "" issue in your 4th/5th gsub calls is almost certainly due to unescaped quotes or incorrect regex syntax. Let's clarify how to handle double quotes in gsub:

If you're trying to target consecutive double quotes, use single quotes to wrap your regex (so you don't have to escape the double quotes inside):

# Replace consecutive double quotes with a single one
fixed_quotes <- gsub('""', '"', cleaned_encoding)

# Or remove all double quotes entirely
no_quotes <- gsub('"', '', cleaned_encoding)

# If you need to match a pattern like ""text"", use this to extract the inner text
extracted_text <- gsub('""([^"]*)""', '\\1', cleaned_encoding)

If you prefer using double quotes for your regex, you'll need to escape the actual double quote characters with \":

fixed_quotes <- gsub("\"\"", "\"", cleaned_encoding)

3. Full Workflow Example (Batch Docs to Excel) #

Here's a complete script that combines encoding fixes, quote cleaning, and batch processing to export to Excel:

library(officer)
library(writexl)
library(dplyr)

# Get all Word docs in a folder
doc_folder <- "path/to/your/docs"
doc_paths <- list.files(doc_folder, pattern = "\\.docx$", full.names = TRUE)

# Process each doc
processed_data <- lapply(doc_paths, function(path) {
  # Read doc content
  doc <- read_docx(path)
  text <- docx_summary(doc)$content
  
  # Fix encoding
  text <- iconv(text, from = "latin1", to = "UTF-8", sub = "?")
  
  # Clean up quotes and junk characters
  text <- gsub('""', '"', text) # Fix consecutive quotes
  text <- gsub("[^[:print:]]", "", text) # Remove unprintable chars
  text <- trimws(text) # Trim extra whitespace
  
  # Return as a data frame row
  data.frame(Document_Name = basename(path), Extracted_Text = text)
})

# Combine all results and export to Excel
final_output <- bind_rows(processed_data)
write_xlsx(final_output, "extracted_text_results.xlsx")

4. Quick Troubleshooting Tips #

• If encoding errors persist, check the actual encoding of your text with stringi::stri_enc_detect(raw_text)—use the detected encoding as the from value in iconv.
• If your "double quotes" are actually full-width (like “” instead of ""), adjust your gsub to target those specific characters: gsub("“”", "\"", text)

内容的提问来源于stack exchange，提问作者jonvas