Consistently strip trailing \n when parsing

[hadley]

I can't tell what the original logic was supposed to be, and it seemed inconsistent. Given that the newline was sometimes present and sometimes absent, I assume downstream code must already deal with the differences so this change won't break anything, but we'll need to double-check that. I have confirmed that all testthat tests still pass and it doesn't affect any snapshot output.

Includes some code quality improvements.

[hadley]

I've gone a bit further with this refactoring, but I have been careful to not further affect the algorithm, just improve the variable names so it's a bit easier to understand what's happening. I've also added quite a few comments to help me understand what is going on.

[cderv]

I don't know exactly the logic - we would need to dig into commit history and related issues to understand. I know newlines in chunk output for knitr is something sensitive

Though, this PR is currently breaking some knitr examples tests where newlines are expected

• https://github.com/yihui/knitr/actions/runs/9659770135/job/26643862541#step:12:1025

Diff patch of examples' changes

diff --git a/007-text-output.md b/007-text-output.md
index 38f9128..92e7335 100644
--- a/007-text-output.md
+++ b/007-text-output.md
@@ -27,8 +27,6 @@ for (i in 1:10) {
 
 ``` r
 # two blank lines below
-
-
 dnorm(0)
 ```
 
@@ -77,8 +75,6 @@ dnorm(0)
 
 ``` r
 > # two blank lines below
-> 
-> 
 > dnorm(0)
 ```
 
@@ -174,8 +170,6 @@ for (i in 1:10) {
 
 ``` r
 # two blank lines below
-
-
 dnorm(0)
 ```
 
@@ -200,8 +194,6 @@ dnorm(0)
 
 ``` r
 # two blank lines below
-
-
 dnorm(0)
 ```
 
@@ -328,8 +320,6 @@ R> for (i in 1:10) {
 
 ``` r
 R> # two blank lines below
-R> 
-R> 
 R> dnorm(0)
 ```
 
diff --git a/051-flowchart.tex b/051-flowchart.tex
index 999a4cc..09cbbe5 100644
--- a/051-flowchart.tex
+++ b/051-flowchart.tex
@@ -66,7 +66,6 @@
 \begin{alltt}
 \hlcom{## Flowchart examples}
 \hlkwd{par}\hldef{(}\hlkwc{ask} \hldef{=} \hlnum{TRUE}\hldef{)}
-
 \hlcom{## MODELLING DIAGRAM}
 \hldef{mar} \hlkwb{<-} \hlkwd{par}\hldef{(}\hlkwc{mar} \hldef{=} \hlkwd{c}\hldef{(}\hlnum{1}\hldef{,} \hlnum{1}\hldef{,} \hlnum{1}\hldef{,} \hlnum{1}\hldef{))}
 \hlkwd{openplotmat}\hldef{(}\hlkwc{main} \hldef{=} \hlsng{"from Soetaert and herman, book in prep"}\hldef{,}
@@ -78,29 +77,22 @@
     \hlkwc{arr.side} \hldef{=} \hlnum{3}\hldef{,} \hlkwc{endhead} \hldef{=} \hlnum{TRUE}\hldef{)}
 \hlkwd{segmentarrow}\hldef{(elpos[}\hlnum{7}\hldef{, ], elpos[}\hlnum{4}\hldef{, ],} \hlkwc{arr.pos} \hldef{=} \hlnum{0.15}\hldef{,} \hlkwc{dd} \hldef{=} \hlnum{0.3}\hldef{,}
     \hlkwc{arr.side} \hldef{=} \hlnum{3}\hldef{,} \hlkwc{endhead} \hldef{=} \hlnum{TRUE}\hldef{)}
-
 \hldef{pin} \hlkwb{<-} \hlkwd{par}\hldef{(}\hlsng{"pin"}\hldef{)}  \hlcom{# size of plotting region, inches}
 \hldef{xx} \hlkwb{<-} \hlnum{0.2}
 \hldef{yy} \hlkwb{<-} \hldef{xx} \hlopt{*} \hldef{pin[}\hlnum{1}\hldef{]}\hlopt{/}\hldef{pin[}\hlnum{2}\hldef{]} \hlopt{*} \hlnum{0.15}  \hlcom{# used to make circles round}
-
 \hldef{sx} \hlkwb{<-} \hlkwd{rep}\hldef{(xx,} \hlnum{8}\hldef{)}
 \hldef{sx[}\hlnum{7}\hldef{]} \hlkwb{<-} \hlnum{0.05}
-
 \hldef{sy} \hlkwb{<-} \hlkwd{rep}\hldef{(yy,} \hlnum{8}\hldef{)}
 \hldef{sy[}\hlnum{6}\hldef{]} \hlkwb{<-} \hldef{yy} \hlopt{*} \hlnum{1.5}
 \hldef{sy[}\hlnum{7}\hldef{]} \hlkwb{<-} \hldef{sx[}\hlnum{7}\hldef{]} \hlopt{*} \hldef{pin[}\hlnum{1}\hldef{]}\hlopt{/}\hldef{pin[}\hlnum{2}\hldef{]}
-
 \hlkwa{for} \hldef{(i} \hlkwa{in} \hlkwd{c}\hldef{(}\hlnum{1}\hlopt{:}\hlnum{7}\hldef{))} \hlkwd{straightarrow}\hldef{(}\hlkwc{to} \hldef{= elpos[i} \hlopt{+} \hlnum{1}\hldef{, ],} \hlkwc{from} \hldef{= elpos[i,}
     \hldef{],} \hlkwc{lwd} \hldef{=} \hlnum{2}\hldef{,} \hlkwc{arr.pos} \hldef{=} \hlnum{0.6}\hldef{,} \hlkwc{endhead} \hldef{=} \hlnum{TRUE}\hldef{)}
 \hldef{lab} \hlkwb{<-} \hlkwd{c}\hldef{(}\hlsng{"Problem"}\hldef{,} \hlsng{"Conceptual model"}\hldef{,} \hlsng{"Mathematical model"}\hldef{,}
     \hlsng{"Parameterisation"}\hldef{,} \hlsng{"Mathematical solution"}\hldef{,} \hlsng{""}\hldef{,} \hlsng{"OK?"}\hldef{,} \hlsng{"Prediction, Analysis"}\hldef{)}
-
 \hlkwa{for} \hldef{(i} \hlkwa{in} \hlkwd{c}\hldef{(}\hlnum{1}\hlopt{:}\hlnum{5}\hldef{,} \hlnum{8}\hldef{))} \hlkwd{textround}\hldef{(elpos[i, ], sx[i], sy[i],} \hlkwc{lab} \hldef{= lab[i])}
-
 \hlkwd{textround}\hldef{(elpos[}\hlnum{6}\hldef{, ], xx, yy} \hlopt{*} \hlnum{1.5}\hldef{,} \hlkwc{lab} \hldef{=} \hlkwd{c}\hldef{(}\hlsng{"Calibration,sensitivity"}\hldef{,}
     \hlsng{"Verification,validation"}\hldef{))}
 \hlkwd{textdiamond}\hldef{(elpos[}\hlnum{7}\hldef{, ], sx[}\hlnum{7}\hldef{], sy[}\hlnum{7}\hldef{],} \hlkwc{lab} \hldef{= lab[}\hlnum{7}\hldef{])}
-
 \hlkwd{textplain}\hldef{(}\hlkwd{c}\hldef{(}\hlnum{0.7}\hldef{, elpos[}\hlnum{2}\hldef{,} \hlnum{2}\hldef{]), yy} \hlopt{*} \hlnum{2}\hldef{,} \hlkwc{lab} \hldef{=} \hlkwd{c}\hldef{(}\hlsng{"main components"}\hldef{,}
     \hlsng{"relationships"}\hldef{),} \hlkwc{font} \hldef{=} \hlnum{3}\hldef{,} \hlkwc{adj} \hldef{=} \hlkwd{c}\hldef{(}\hlnum{0}\hldef{,} \hlnum{0.5}\hldef{))}
 \hlkwd{textplain}\hldef{(}\hlkwd{c}\hldef{(}\hlnum{0.7}\hldef{, elpos[}\hlnum{3}\hldef{,} \hlnum{2}\hldef{]), yy,} \hlsng{"general theory"}\hldef{,} \hlkwc{adj} \hldef{=} \hlkwd{c}\hldef{(}\hlnum{0}\hldef{,}
@@ -114,7 +106,6 @@
 \includegraphics[width=\maxwidth]{figure/051-flowchart-demo-flowchart-1} 
 \begin{kframe}\begin{alltt}
 \hlcom{##### DIAGRAM}
-
 \hlkwd{par}\hldef{(}\hlkwc{mar} \hldef{=} \hlkwd{c}\hldef{(}\hlnum{1}\hldef{,} \hlnum{1}\hldef{,} \hlnum{1}\hldef{,} \hlnum{1}\hldef{))}
 \hlkwd{openplotmat}\hldef{()}
 \hldef{elpos} \hlkwb{<-} \hlkwd{coordinates}\hldef{(}\hlkwd{c}\hldef{(}\hlnum{1}\hldef{,} \hlnum{1}\hldef{,} \hlnum{2}\hldef{,} \hlnum{4}\hldef{))}
@@ -141,7 +132,6 @@
 \hlkwd{textellipse}\hldef{(elpos[}\hlnum{8}\hldef{, ],} \hlnum{0.1}\hldef{,} \hlnum{0.1}\hldef{,} \hlkwc{lab} \hldef{=} \hlkwd{c}\hldef{(}\hlsng{"new"}\hldef{,} \hlsng{"article"}\hldef{),}
     \hlkwc{box.col} \hldef{=} \hlsng{"orange"}\hldef{,} \hlkwc{shadow.col} \hldef{=} \hlsng{"red"}\hldef{,} \hlkwc{shadow.size} \hldef{=} \hlnum{0.005}\hldef{,}
     \hlkwc{cex} \hldef{=} \hlnum{1.5}\hldef{)}
-
 \hldef{dd} \hlkwb{<-} \hlkwd{c}\hldef{(}\hlnum{0}\hldef{,} \hlnum{0.025}\hldef{)}
 \hlkwd{text}\hldef{(arrpos[}\hlnum{2}\hldef{,} \hlnum{1}\hldef{]} \hlopt{+} \hlnum{0.05}\hldef{, arrpos[}\hlnum{2}\hldef{,} \hlnum{2}\hldef{],} \hlsng{"yes"}\hldef{)}
 \hlkwd{text}\hldef{(arrpos[}\hlnum{3}\hldef{,} \hlnum{1}\hldef{]} \hlopt{-} \hlnum{0.05}\hldef{, arrpos[}\hlnum{3}\hldef{,} \hlnum{2}\hldef{],} \hlsng{"no"}\hldef{)}
@@ -159,19 +149,16 @@
 \hlkwd{treearrow}\hldef{(}\hlkwc{from} \hldef{= elpos[}\hlnum{1}\hlopt{:}\hlnum{2}\hldef{, ],} \hlkwc{to} \hldef{= elpos[}\hlnum{3}\hlopt{:}\hlnum{5}\hldef{, ],} \hlkwc{arr.side} \hldef{=} \hlnum{2}\hldef{,}
     \hlkwc{path} \hldef{=} \hlsng{"H"}\hldef{)}
 \hlkwa{for} \hldef{(i} \hlkwa{in} \hlnum{1}\hlopt{:}\hlnum{5}\hldef{)} \hlkwd{textrect}\hldef{(elpos[i, ],} \hlnum{0.15}\hldef{,} \hlnum{0.05}\hldef{,} \hlkwc{lab} \hldef{= i,} \hlkwc{cex} \hldef{=} \hlnum{1.5}\hldef{)}
-
 \hlkwd{openplotmat}\hldef{()}
 \hldef{elpos} \hlkwb{<-} \hlkwd{coordinates}\hldef{(}\hlkwd{c}\hldef{(}\hlnum{3}\hldef{,} \hlnum{2}\hldef{),} \hlkwc{hor} \hldef{=} \hlnum{FALSE}\hldef{)}
 \hlkwd{treearrow}\hldef{(}\hlkwc{from} \hldef{= elpos[}\hlnum{1}\hlopt{:}\hlnum{3}\hldef{, ],} \hlkwc{to} \hldef{= elpos[}\hlnum{4}\hlopt{:}\hlnum{5}\hldef{, ],} \hlkwc{arr.side} \hldef{=} \hlnum{2}\hldef{,}
     \hlkwc{arr.pos} \hldef{=} \hlnum{0.2}\hldef{,} \hlkwc{path} \hldef{=} \hlsng{"V"}\hldef{)}
 \hlkwa{for} \hldef{(i} \hlkwa{in} \hlnum{1}\hlopt{:}\hlnum{5}\hldef{)} \hlkwd{textrect}\hldef{(elpos[i, ],} \hlnum{0.15}\hldef{,} \hlnum{0.05}\hldef{,} \hlkwc{lab} \hldef{= i,} \hlkwc{cex} \hldef{=} \hlnum{1.5}\hldef{)}
-
 \hlkwd{openplotmat}\hldef{()}
 \hldef{elpos} \hlkwb{<-} \hlkwd{coordinates}\hldef{(}\hlkwd{c}\hldef{(}\hlnum{1}\hldef{,} \hlnum{4}\hldef{))}
 \hlkwd{treearrow}\hldef{(}\hlkwc{from} \hldef{= elpos[}\hlnum{1}\hldef{, ],} \hlkwc{to} \hldef{= elpos[}\hlnum{2}\hlopt{:}\hlnum{5}\hldef{, ],} \hlkwc{arr.side} \hldef{=} \hlnum{2}\hldef{,}
     \hlkwc{arr.pos} \hldef{=} \hlnum{0.7}\hldef{,} \hlkwc{path} \hldef{=} \hlsng{"H"}\hldef{)}
 \hlkwa{for} \hldef{(i} \hlkwa{in} \hlnum{1}\hlopt{:}\hlnum{5}\hldef{)} \hlkwd{textrect}\hldef{(elpos[i, ],} \hlnum{0.05}\hldef{,} \hlnum{0.05}\hldef{,} \hlkwc{lab} \hldef{= i,} \hlkwc{cex} \hldef{=} \hlnum{1.5}\hldef{)}
-
 \hlkwd{openplotmat}\hldef{()}
 \hldef{elpos} \hlkwb{<-} \hlkwd{coordinates}\hldef{(}\hlkwd{c}\hldef{(}\hlnum{2}\hldef{,} \hlnum{1}\hldef{,} \hlnum{2}\hldef{,} \hlnum{3}\hldef{))}
 \hldef{elpos[}\hlnum{1}\hldef{,} \hlnum{1}\hldef{]} \hlkwb{<-} \hlnum{0.3}
@@ -188,7 +175,6 @@
 \includegraphics[width=\maxwidth]{figure/051-flowchart-demo-flowchart-3} 
 \begin{kframe}\begin{alltt}
 \hlkwd{par}\hldef{(}\hlkwc{mfrow} \hldef{=} \hlkwd{c}\hldef{(}\hlnum{1}\hldef{,} \hlnum{1}\hldef{))}
-
 \hlkwd{par}\hldef{(}\hlkwc{mar} \hldef{=} \hlkwd{c}\hldef{(}\hlnum{0}\hldef{,} \hlnum{0}\hldef{,} \hlnum{0}\hldef{,} \hlnum{0}\hldef{))}
 \hlkwd{openplotmat}\hldef{()}
 \hldef{elpos} \hlkwb{<-} \hlkwd{coordinates}\hldef{(}\hlkwd{c}\hldef{(}\hlnum{1}\hldef{,} \hlnum{1}\hldef{,} \hlnum{2}\hldef{,} \hlnum{1}\hldef{))}
@@ -229,9 +215,7 @@
     \hlkwc{lcol} \hldef{=} \hlnum{6}\hldef{)}
 \hlkwd{splitarrow}\hldef{(}\hlkwc{from} \hldef{= elpos[}\hlnum{1}\hldef{, ],} \hlkwc{to} \hldef{= elpos[}\hlnum{2}\hlopt{:}\hlnum{3}\hldef{, ],} \hlkwc{lty} \hldef{=} \hlnum{1}\hldef{,} \hlkwc{lwd} \hldef{=} \hlnum{1}\hldef{,}
     \hlkwc{dd} \hldef{=} \hlnum{0.7}\hldef{,} \hlkwc{arr.side} \hldef{=} \hlnum{1}\hlopt{:}\hlnum{2}\hldef{,} \hlkwc{lcol} \hldef{=} \hlnum{7}\hldef{)}
-
 \hlkwa{for} \hldef{(i} \hlkwa{in} \hlnum{1}\hlopt{:}\hlnum{4}\hldef{)} \hlkwd{textrect}\hldef{(elpos[i, ],} \hlnum{0.05}\hldef{,} \hlnum{0.05}\hldef{,} \hlkwc{lab} \hldef{= i,} \hlkwc{cex} \hldef{=} \hlnum{1.5}\hldef{)}
-
 \hlkwd{legend}\hldef{(}\hlsng{"topright"}\hldef{,} \hlkwc{lty} \hldef{=} \hlnum{1}\hlopt{:}\hlnum{7}\hldef{,} \hlkwc{legend} \hldef{=} \hlkwd{c}\hldef{(}\hlsng{"segmentarrow"}\hldef{,} \hlsng{"curvedarrow"}\hldef{,}
     \hlsng{"straightarrow"}\hldef{,} \hlsng{"treearrow"}\hldef{,} \hlsng{"bentarrow"}\hldef{,} \hlsng{"selfarrow"}\hldef{,} \hlsng{"splitarrow"}\hldef{),}
     \hlkwc{lwd} \hldef{=} \hlkwd{c}\hldef{(}\hlkwd{rep}\hldef{(}\hlnum{2}\hldef{,} \hlnum{6}\hldef{),} \hlnum{1}\hldef{),} \hlkwc{col} \hldef{=} \hlnum{1}\hlopt{:}\hlnum{7}\hldef{)}
diff --git a/093-knitr-asy.md b/093-knitr-asy.md
index 18136f7..a84d135 100644
--- a/093-knitr-asy.md
+++ b/093-knitr-asy.md
@@ -62,7 +62,6 @@ In this example, I  generate data in R and then use it for plotting in asymptote
 ``` r
 x = seq(0, 5, l = 100)
 y = sin(x)
-
 # save data to csv file
 write.table(data.frame(x, y), file = "asy.csv", col.names = FALSE, row.names = FALSE,
     sep = ",")

However, I can't reproduce it outside of this context, so I need to check what is happening. I can reproduce it locally with knitr example, but not running evaluate directly

[cderv]

In fact, I have this change in behavior which is expected by this PR I guess

With CRAN evaluate

> str(evaluate::evaluate(input = "# two blank lines below\n\n\ndnorm(0)"))
List of 5
 $ :List of 1
  ..$ src: chr "# two blank lines below\n"
  ..- attr(*, "class")= chr "source"
 $ :List of 1
  ..$ src: chr "\n"
  ..- attr(*, "class")= chr "source"
 $ :List of 1
  ..$ src: chr "\n"
  ..- attr(*, "class")= chr "source"
 $ :List of 1
  ..$ src: chr "dnorm(0)"
  ..- attr(*, "class")= chr "source"
 $ : chr "[1] 0.3989423\n"

With this PR version

> str(evaluate(input = "# two blank lines below\n\n\ndnorm(0)"))
List of 5
 $ :List of 1
  ..$ src: chr "# two blank lines below"
  ..- attr(*, "class")= chr "source"
 $ :List of 1
  ..$ src: chr ""
  ..- attr(*, "class")= chr "source"
 $ :List of 1
  ..$ src: chr ""
  ..- attr(*, "class")= chr "source"
 $ :List of 1
  ..$ src: chr "dnorm(0)"
  ..- attr(*, "class")= chr "source"
 $ : chr "[1] 0.3989423\n"
 - attr(*, "class")= chr [1:2] "evaluate_evaluation" "list"

But it seems it will impact knitr which does not seem to add the new lines back everywhere 🤔

Still need to find where

[hadley]

Hmmm, if it's going to break knitr I probably won't merge it. Maybe the rule is that evaluate adds a \n whenever the source would otherwise be blank?

[hadley]

Replaced by #183