如何讀取復雜的 .txt 文件並轉換為 JSON

Question

我需要從我們的 MRT 掃描儀生成的 txt 文件中提取信息，然后將信息轉換為具有特定結構的 JSON 文件。 該文本文件不是數據文件，而是包含有關該掃描會話的信息的實際文本文件。 任何人都可以讓我走上正確的道路嗎？

下面是部分文件內容的示例。 我更喜歡用 R 來做，但 MATLAB 或 Python 也是可能的。

Example:

Image filter =          "system default";
Uniformity correction =     "no";
Geometry correction =       "default";
IF_info_seperator =     0;
Total scan duration =       "09:05.0";
Rel. SNR =          0.752056241;
Act. TR (ms) =          "5000";
Act. TE (ms) =          "74";
ACQ matrix M x P =      "96 x 94";
ACQ voxel MPS (mm) =        "2.50 / 2.55 / 2.50";
REC voxel MPS (mm) =        "2.50 / 2.50 / 2.50";
Scan percentage (%) =       97.9166641;

Answer 1

在 R 中，使用readLines 、 gsub和同事。

## read raw lines; don't warn "final line" (might have unwanted side-effects) 
txt <- readLines("mri.txt", warn=FALSE)
## replace first `=` with `§` and split there
spl <- strsplit(sub("=", "§", txt), "§")
## expand lengths of list elements to 2 (throws a `NA` in case)
spl <- lapply(spl, `length<-`, 2)
## remove leading/trailing whitespace; make matrix
tmp <- t(sapply(spl, trimws))
## replace `;` or `"` with empty string,
tmp[,2] <- gsub(";|\"", "", tmp[,2])
tmp
#       [,1]                    [,2]                
#  [1,] "Image filter"          "system default"    
#  [2,] "Uniformity correction" "no"                
#  [3,] "Geometry correction"   "default"           
#  [4,] "IF_info_seperator"     "0"                 
#  [5,] "Total scan duration"   "09:05.0"           
#  [6,] "Rel. SNR"              "0.752056241"       
#  [7,] "Act. TR (ms)"          "5000"              
#  [8,] "Act. TE (ms)"          "74"                
#  [9,] "ACQ matrix M x P"      "96 x 94"           
# [10,] "ACQ voxel MPS (mm)"    "2.50 / 2.55 / 2.50"
# [11,] "REC voxel MPS (mm)"    "2.50 / 2.50 / 2.50"
# [12,] "Scan percentage (%)"   "97.9166641"   

最后是jsonlite::toJSON 。

library(jsonlite)
toJSON(tmp)
[["Image filter","system default"],["Uniformity correction","no"],["Geometry correction","default"],["IF_info_seperator","0"],["Total scan duration","09:05.0"],["Rel. SNR","0.752056241"],["Act. TR (ms)","5000"],["Act. TE (ms)","74"],["ACQ matrix M x P","96 x 94"],["ACQ voxel MPS (mm)","2.50 / 2.55 / 2.50"],["REC voxel MPS (mm)","2.50 / 2.50 / 2.50"],["Scan percentage (%)","97.9166641"]] 

當然，您可能需要對此進行微調。