R SQLove 用于复杂查询的 ODBC 连接

Question

我正在尝试将复杂查询的结果放入 r 中进行操作/ETL。我找到了 SQLove，听起来它正是能让我的生活变得轻松的软件包。尝试在一个简单的脚本上测试它 -

if (!require("pacman")
) install.packages("pacman")

pacman::p_load(
  #add list of libraries here
  SQLove,
  odbc,
  RJDBC
)

con <- 
  dbConnect(
    odbc::odbc(),
    driver = "SQL Server",
    server = "ClarDbPrd_Alias",
    database = "Clarity"
  )

d1 <- 
  dbGetMultiQuery(
    connection = con,
    sql_file_path = "S:/QITeam/DataAnalytics/Projects/Frailty/Scripts/test.sql"
    )

SQL 文件是一个简单的 Select * 来测试包是否有效

SELECT * FROM CLARITY_DEP;

当我运行它时，出现以下错误：

Error: unable to find an inherited method for function ‘dbSendUpdate’ for signature ‘conn = "Microsoft SQL Server", statement = "SQL"’

尝试研究这个问题，看起来 RJDBC 包中的 dbSendUpdate 仅适用于 JBDC，不适用于 ODBC。是否有解决方法或不同的包可以轻松运行 ODBC .sql 文件类型中的脚本？我喜欢它说它清理评论和所有内容的部分（两者 - 和多行 /**/）

https://samkerns.r-universe.dev/articles/SQLove/SQLove.html

Answer 1

在 GitHub 上对该函数进行一些研究后，我认为它已经过时了。

https://github.com/samkerns/SQLove/blob/main/R/dbGetMultiQuery.R

所以我拿起代码，根据我的目的对其进行了一些调整，并且在我用于工作的更简单的 SQL 脚本上，它毫无问题地弹出了结果

# Check to see if pacman is installed
## Install if it is not loaded
if (!require("pacman")
) install.packages("pacman")

# Get needed libraries
pacman::p_load(
  odbc,
  DBI,
  stringr
)

# Set up connection
con <-
  dbConnect(
    odbc::odbc(),
    driver = "SQL Server",
    server = "ClarDbPrd_Alias",
    database = "Clarity"
  )

# Function to read in SQL file execute all commands and return final query
dbGetMultiQuery <- 
  function(
    connection, # connection set up from dbConnect
    sql_file_path # path to the SQL file
    ) {
    
    #Reading in the SQL file
    sql_file <-
      readr::read_file(sql_file_path)
    
    #Removing all comments /* and --
    ## /**/ comments
    sql_file <- base::gsub("/\\*.*?\\*/", "", sql_file)
    
    ## -- comments
    sql_file <- base::gsub("--.*?\\r", "\\\r", sql_file)
    
    ## special characters
    sql_file <- str_replace_all(sql_file, "[\r\n]", "  ")
    sql_file <- str_replace_all(sql_file, "[\t]", " ")

    ## white space before and after to make debugging easier
    sql_file <- trimws(sql_file)
    
    # Splitting the SQL file into individual queries
    ## ignore '; ' because I use that in STRING_AGG
    sql_list <- base::strsplit(sql_file, "(?<!')\\;(?! ')", perl = TRUE)
    
    # Get the number of queries
    query_length <- base::lengths(sql_list)
    
    #If only 1 query is available, it's a SELECT statement, use DBI::dbGetQuery
    if (query_length == 1) {
      DBI::dbGetQuery(con, sql_list[[1]][[1]])
      
      #If more than 1 query is available, dbExecute for all but final statement
      } else {
        for (i in c(1:(query_length - 1))) {
          DBI::dbExecute(con, DBI::SQL(sql_list[[1]][[i]]))
          print(paste("Statement", i, "of", query_length, "complete"))
          }
        #Create dataframe from final query statement
        DBI::dbGetQuery(con, sql_list[[1]][[query_length]])
      }
    }

# Test
df1 <- 
  dbGetMultiQuery(
    connection = con, 
    sql_file_path = "S:/QITeam/DataAnalytics/Projects/Frailty/Scripts/CasesGeriatric_1c.sql"
    )

我希望人们在希望在 R 中执行多个查询时遇到这个问题，因为我希望当我第一次开始搜索如何在 R 中执行此操作时遇到这样的答案:)