$ cat fields.txt
xxxxxxxxxxxxxxxxxxxxxxxxxxxxx  yyyyyyyyyyyyyyyyyyyyyyyyyyyy 111111
xxxxxxxxxxxxxxxxxxxxxxxxxxx    yyyyyyyyyyyyyyyyyyyyyyyy     222222
xxxxxxxxxxxxxxxxxxxxxxxxxxxx  yyyyyyyyyyyyyyyyyyyyyyyyyyyy 3333333
xxxxxxxxxxxxxxxxxxxxxxxxxx    yyyyyyyyyyyyyyyyyyyyyyyy     4444444

# wrong output for 2nd and 4th lines, and the failure is different
$ frawk '{print NR ":" $3 }' fields.txt
1:111111
2:yyyyyyyyyyyyyyyyyyyyyyyy
3:3333333
4:

# works correctly if those lines are given as the sole input
$ sed -n '2p' fields.txt | frawk '{print NR ":" $3 }'
1:222222
$ sed -n '4p' fields.txt | frawk '{print NR ":" $3 }'
1:4444444

The failure also seems to depend on the length of the input lines or something like that, which is why I have those long x and y in the input, couldn't find a simpler failing case.

Hi, I'm not a rust programmer so I might be doing something wrong here but I couldn't compile frawk.

I tried following ways to install

cargo +nightly install --path .
cargo  install frawk
cargo +nightly install --path . --no-default-features --features use_jemalloc,allow_avx2,unstable
cargo +nightly install --path . --no-default-features --features allow_avx2,unstable

all failed with the following message:

error[E0599]: no method named `multiple` found for struct `clap::Arg` in the current scope
   --> src/main.rs:306:15
    |
306 |              .multiple(true)
    |               ^^^^^^^^ method not found in `clap::Arg<'_>`

error[E0599]: no method named `multiple` found for struct `clap::Arg` in the current scope
   --> src/main.rs:329:15
    |
329 |              .multiple(true)
    |               ^^^^^^^^ method not found in `clap::Arg<'_>`

error[E0599]: no method named `multiple` found for struct `clap::Arg` in the current scope
   --> src/main.rs:350:15
    |
350 |              .multiple(true))
    |               ^^^^^^^^ method not found in `clap::Arg<'_>`

For more information about this error, try `rustc --explain E0599`.
warning: `frawk` (bin "frawk") generated 2 warnings
error: failed to compile `frawk v0.4.2 (/usr/src/myapp)`, intermediate artifacts can be found at `/usr/src/myapp`

I'm using ubuntu 20.04 and I ran rustup default nightly for nightly compilation. Here are the version numbers in case they are needed:

# cargo --version
cargo 1.56.0-nightly (e96bdb0c3 2021-08-17)
# rustc --version
rustc 1.56.0-nightly (ad02dc46b 2021-08-26)
# rustup --version
rustup 1.24.3 (ce5817a94 2021-05-31)
info: This is the version for the rustup toolchain manager, not the rustc compiler.
info: The currently active `rustc` version is `rustc 1.56.0-nightly (ad02dc46b 2021-08-26)`

Given foobar.csv:

foo,bar
1,2
3,4

I'd like to be able to write bar_mean.awk:

{ acc += FIELDS["bar"] }
END { print "bar's mean is " acc/NR }

and run it like this:

$ frawk -icsv --headers -f bar_mean.awk <foobar.csv
bar's mean is 3

Differences when --headers is passed:

• The first line is skipped: it's neither processed by the awk program, nor does it count towards NR.
• A magic hashmap is defined (I called it FIELDS for the sake of the example above) which maps header name to field value.

If you're open to going crazy with the syntax, you could even do something like $"bar". (I think $bar is probably going too far, since that's already valid AWK.)

I am trying to build frawk without LLVM (#67) and am getting:

$ cargo +nightly install --path . --no-default-features --features use_jemalloc,allow_avx2,unstable
   Compiling cranelift-jit v0.75.0
error[E0034]: multiple applicable items in scope
  --> /home/motiejus/.cargo/registry/src/github.com-1ecc6299db9ec823/lalrpop-0.17.2/src/message/horiz.rs:22:14
   |
22 |             .intersperse(self.separate)
   |              ^^^^^^^^^^^ multiple `intersperse` found
   |
   = note: candidate #1 is defined in an impl of the trait `Iterator` for the type `std::iter::Map<I, F>`
   = note: candidate #2 is defined in an impl of the trait `Itertools` for the type `T`
help: disambiguate the associated function for candidate #1
   |
19 ~         Iterator::intersperse(self.items
20 +             .iter()
21 +             .map(|c| c.min_width()), self.separate)
   |
help: disambiguate the associated function for candidate #2
   |
19 ~         Itertools::intersperse(self.items
20 +             .iter()
21 +             .map(|c| c.min_width()), self.separate)

I recognize it's a broken dependency and should be reported upstream. However, I looked upstream, but didn't find a way to compile the standalone project and report it there. Guidance would be appreciated; I am new to rust ecosystem.

motiejus ~/code/frawk $ rustup show
Default host: x86_64-unknown-linux-gnu
rustup home:  /home/motiejus/.rustup

nightly-x86_64-unknown-linux-gnu (default)
rustc 1.57.0-nightly (fdf65053e 2021-09-07)
motiejus ~/code/frawk $ rustc --version
rustc 1.57.0-nightly (fdf65053e 2021-09-07)
motiejus ~/code/frawk $ cargo --version
cargo 1.56.0-nightly (18751dd3f 2021-09-01)
motiejus ~/code/frawk $ 

frawk does not detect start of program without -- when -v is last argument before program start.

❯ seq 1 10 | frawk -v multiplier=20 '{ print $0 * multiplier }'
must specify program at command line, or in a file via -f

❯ seq 1 10 | frawk -v multiplier=20 -- '{ print $0 * multiplier }'
20.0
40.0
60.0
80.0
100.0
120.0
140.0
160.0
180.0
200.0

❯ seq 1 10 | frawk -F '\t' -v multiplier=20 '{ print $0 * multiplier }'
must specify program at command line, or in a file via -f

~/Downloads via ☕ v1.8.0 via 🐍 v3.6.7
❯ seq 1 10 | frawk -v multiplier=20 -F '\t' '{ print $0 * multiplier }'
20.0
40.0
60.0
80.0
100.0
120.0
140.0
160.0
180.0
200.0

frawk also seems to have problems with parsing variables with dashes (and maybe other special characters):

❯ seq 1 10 | awk -v multiplier=20 -v var_with_dash=variable-with-dash -F '\t' '{ print $0 * multiplier, var_with_dash }'
20 variable-with-dash
40 variable-with-dash
60 variable-with-dash
80 variable-with-dash
100 variable-with-dash
120 variable-with-dash
140 variable-with-dash
160 variable-with-dash
180 variable-with-dash
200 variable-with-dash

❯ seq 1 10 | frawk -v multiplier=20 -v var_with_dash=variable-with-dash -F '\t' '{ print $0 * multiplier, var_with_dash }'
failed to parse var at index 2:
var_with_dash=variable-with-dash
error:UnrecognizedToken { token: (Loc { line: 0, col: 22, offset: 22 }, Sub, Loc { line: 0, col: 23, offset: 23 }), expected: ["\"[\""] }

❯ seq 1 10 | frawk -v multiplier=20 -v 'var_with_dash=variable-with-dash' -F '\t' '{ print $0 * multiplier, var_with_dash }'
failed to parse var at index 2:
var_with_dash=variable-with-dash
error:UnrecognizedToken { token: (Loc { line: 0, col: 22, offset: 22 }, Sub, Loc { line: 0, col: 23, offset: 23 }), expected: ["\"[\""] }

❯ seq 1 10 | frawk -v multiplier=20 -v 'var_with_dash="variable-with-dash"' -F '\t' '{ print $0 * multiplier, var_with_dash }'
20.0 variable-with-dash
40.0 variable-with-dash
60.0 variable-with-dash
80.0 variable-with-dash
100.0 variable-with-dash
120.0 variable-with-dash
140.0 variable-with-dash
160.0 variable-with-dash
180.0 variable-with-dash
200.0 variable-with-dash

❯ seq 1 10 | frawk -v multiplier=20 -v var_with_dash="variable-with-dash" -F '\t' '{ print $0 * multiplier, var_with_dash }'
failed to parse var at index 2:
var_with_dash=variable-with-dash
error:UnrecognizedToken { token: (Loc { line: 0, col: 22, offset: 22 }, Sub, Loc { line: 0, col: 23, offset: 23 }), expected: ["\"[\""] }

If passed vars are not a number, the variable will be empty if not quoted (once for bash and once for frawk)). Which is quite different from normal awk.

❯ seq 1 10 | frawk -v multiplier=20 -v X=A -v 'var_with_dash="variable-with-dash"' -F '\t' '{ print $0 * multiplier, X, var_with_dash }'
20.0  variable-with-dash
40.0  variable-with-dash
60.0  variable-with-dash
80.0  variable-with-dash
100.0  variable-with-dash
120.0  variable-with-dash
140.0  variable-with-dash
160.0  variable-with-dash
180.0  variable-with-dash
200.0  variable-with-dash

❯ seq 1 10 | frawk -v multiplier=20 -v 'X=A' -v 'var_with_dash="variable-with-dash"' -F '\t' '{ print $0 * multiplier, X, var_with_dash }'
20.0  variable-with-dash
40.0  variable-with-dash
60.0  variable-with-dash
80.0  variable-with-dash
100.0  variable-with-dash
120.0  variable-with-dash
140.0  variable-with-dash
160.0  variable-with-dash
180.0  variable-with-dash
200.0  variable-with-dash

❯ seq 1 10 | frawk -v multiplier=20 -v 'X="A"' -v 'var_with_dash="variable-with-dash"' -F '\t' '{ print $0 * multiplier, X, var_with_dash }'
20.0 A variable-with-dash
40.0 A variable-with-dash
60.0 A variable-with-dash
80.0 A variable-with-dash
100.0 A variable-with-dash
120.0 A variable-with-dash
140.0 A variable-with-dash
160.0 A variable-with-dash
180.0 A variable-with-dash
200.0 A variable-with-dash

This adds support for gensub function, which is a common GNU extension.

This is accomplished by defining it as a built-in function, adding a bytecode instruction for it and finally implementing it as a method on a Str.

I am a bit unsure how to handle runtime errors: should I do something smart or just ad hoc .unwrap()s are fine?

Another thing: not sure whether (and how) I should handle GenSub in feedback function for type propagation

Performance can be improved by handling the very common case of using a constant for "how" parameter and introducing specialized instructions for handling g case (replace all) and number case (replace only n-th match)

Hi,

I'm trying to package frawk for nixos/nixpkgs, but I cannot do that without a Cargo.lock package. Are there any reasons why one isn't included? Thanks.

exit opt_expr An exit statement causes immediate execution of the END actions or program termination if there are none or if the exit occurs in an END action. The opt_expr sets the exit value of the program unless overridden by a later exit or subsequent error.

❯ seq 1 10 | mawk '{ print $0; if ($1 >= 8) { last = $0; exit 33; } } END { print "last line before exit: " last;}' ; echo "Exit code: $?";
1
2
3
4
5
6
7
8
last line before exit: 8
Exit code: 33

❯ seq 1 10 | frawk '{ print $0; if ($1 >= 8) { last = $0; exit 33; } } END { print "last line before exit: " last;}' ; echo "Exit code: $?";
1
2
3
4
5
6
7
8
9
10
last line before exit: 10
Exit code: 0


# Same but with exit with extra parentheses (frawk does not remove extra (), (print had this issue too in the past)):

❯ seq 1 10 | mawk '{ print $0; if ($1 >= 8) { last = $0; exit(33); } } END { print "last line before exit: " last;}' ; echo "Exit code: $?";
1
2
3
4
5
6
7
8
last line before exit: 8
Exit code: 33

❯ seq 1 10 | frawk '{ print $0; if ($1 >= 8) { last = $0; exit(33); } } END { print "last line before exit: " last;}' ; echo "Exit code: $?";
failed to create program context: [src/cfg.rs:1641:21] Call to unknown function "exit"
Exit code: 1

fast-float can be removed when using a recent rust compiler. Similar algoritm now in standard library: https://www.reddit.com/r/rust/comments/omelz4/making_rust_float_parsing_fast_libcore_edition/ https://github.com/pola-rs/polars/issues/1010

When I was running the file info/scripts/groupby.sh I encountered an error at the following line:

time $FRAWK -bllvm -F'\t' -f "$SCRIPT_FILE" "${TSV}"

When I was trying to trıubleshoot the problem I noticed that

cat datafile | frawk -f scriptfile   # works
frawk datafile -f scriptfile         # works
frawk -f scriptfile datafile         # fails

after trying different options, I found out that in case of frawk -f scriptfile datafile the datafile needs to be provided with --. Let me demonstrate the case:

frawk works fine with pipe:

$ echo -e "a,b,c\n1,2,3\n4,5,6"
a,b,c
1,2,3
4,5,6

$ echo -e "a,b,c\n1,2,3\n4,5,6" | frawk -icsv '{ print $2; }'
b
2
5

If we save the script as file, it woks fine in pipe

$ echo '{ print $2; }' > test.awk
$ echo -e "a,b,c\n1,2,3\n4,5,6" | frawk -icsv -f test.awk
b
2
5

if both data and script are from file, then fails

$ echo -e "a,b,c\n1,2,3\n4,5,6" > test.dat
$ frawk -icsv -f test.awk test.dat
Unrecognized token `,` found at line 3, column 4:line 3, column 5
Expected one of "\n", "[" or "{"

workaround is using --

$ frawk -icsv -f test.awk -- test.dat
b
2
5

Not sure if this is bug or a feature but still it would be great if frawk can be drop-in replacement for awk, by simply replacing awk with frawk in scripts , everything should work. So, it would be great if frawk can run just like awk where datafile is provided after script file.

I'm using frawk version 0.4.2 in Linux.

Allow more than one -f argument. Now it results in this error message:

error: The argument '--program-file <FILE>' was provided more than once, but cannot be used multiple times

I have compared nawk with frawk on a macOS Darwin: i7 with 16GB RAM.

I used a 17GB .csv file for test using: frawk 'BEGIN{FS=","};{count[NF]++};END{for(i in count){print "With " i " fields - count is: " count[i]}}' aggregated.csv

results: frawk - 2 min nawk - 23 min

Impressive. Thank for the work on frawk.

This might sound crazy but still I wanted to propose a feature request about parquet files.

You might ask, why? Parquet files are becoming more widespread and might even be considered as "the new csv". There are specialized tools such as duckdb to run sql commands on them. I didn't see or came across any awk-like utility which can process parquet files.

IMHO, supporting parquet files by frawk will be a huge win for "data analysis at the commandline" camp.

yay -Sy frawk .................
Compiling frawk v0.4.6 (/tmp/makepkg/frawk/src/frawk-0.4.6) Compiling tikv-jemallocator v0.4.3 error[E0554]: #![feature] may not be used on the stable release channel --> src/main.rs:2:43 | 2 | #![cfg_attr(feature = "unstable", feature(core_intrinsics))] | ^^^^^^^^^ error[E0554]: #![feature] may not be used on the stable release channel --> src/main.rs:3:43 | 3 | #![cfg_attr(feature = "unstable", feature(test))] | ^^^^ error[E0554]: #![feature] may not be used on the stable release channel --> src/main.rs:4:43 | 4 | #![cfg_attr(feature = "unstable", feature(write_all_vectored))] | ^^^^^^^^^^^^^^^^^^ For more information about this error, try rustc --explain E0554. error: could not compile frawk due to 3 previous errors ==> ERROR: A failure occurred in build().

To my surprise, this crate compiles with cargo build --no-default-features on Windows 10! Woot, awk on Windows! :D

Attempting to compile with the use_jemalloc against x86_64-pc-windows-msvc (the most common Windows target AFAIK), one gets a compile error of the form:

$ cargo check # `default` features enable `use_jemalloc`

<snip>

The following warnings were emitted during compilation:

warning: "jemalloc support for `x86_64-pc-windows-msvc` is untested"

error: failed to run custom build command for `tikv-jemalloc-sys v0.4.2+5.2.1-patched.2`

Caused by:
  process didn't exit successfully: `<snip>\frawk\target\debug\build\tikv-jemalloc-sys-8070626b6b43d50a\build-script-build` (exit code: 101)
  --- stdout
  TARGET=x86_64-pc-windows-msvc
  HOST=x86_64-pc-windows-msvc
  NUM_JOBS=8
  OUT_DIR="<snip>\\target\\debug\\build\\tikv-jemalloc-sys-1f5ef3a81eb3f627\\out"
  BUILD_DIR="<snip>\\target\\debug\\build\\tikv-jemalloc-sys-1f5ef3a81eb3f627\\out\\build"
  SRC_DIR="<snip>\\.cargo\\registry\\src\\github.com-1ecc6299db9ec823\\tikv-jemalloc-sys-0.4.2+5.2.1-patched.2"
  cargo:warning="jemalloc support for `x86_64-pc-windows-msvc` is untested"
  cargo:rustc-cfg=prefixed
  OPT_LEVEL = Some("0")
  TARGET = Some("x86_64-pc-windows-msvc")
  HOST = Some("x86_64-pc-windows-msvc")
  CC_x86_64-pc-windows-msvc = None
  CC_x86_64_pc_windows_msvc = None
  HOST_CC = None
  CC = None
  CFLAGS_x86_64-pc-windows-msvc = None
  CFLAGS_x86_64_pc_windows_msvc = None
  HOST_CFLAGS = None
  CFLAGS = None
  CRATE_CC_NO_DEFAULTS = None
  CARGO_CFG_TARGET_FEATURE = Some("aes,avx,avx2,bmi1,bmi2,fma,fxsr,lzcnt,pclmulqdq,popcnt,rdrand,rdseed,sse,sse2,sse3,sse4.1,sse4.2,ssse3,xsave,xsavec,xsaveopt,xsaves")
  DEBUG = Some("true")
  CC="C:\\Program Files (x86)\\Microsoft Visual Studio\\2017\\Community\\VC\\Tools\\MSVC\\14.12.25827\\bin\\HostX64\\x64\\cl.exe"
  CFLAGS="-nologo -MD -Z7 -Brepro -W4"
  JEMALLOC_REPO_DIR="jemalloc"
  --with-jemalloc-prefix=_rjem_
  running: "sh" "<snip>/frawk/target/debug/build/tikv-jemalloc-sys-1f5ef3a81eb3f627/out/build/configure" "--disable-cxx" "--with-jemalloc-prefix=_rjem_" "--with-private-namespace=_rjem_" "--host=x86_64-pc-win32" "--build=x86_64-pc-win32" "--prefix=<snip>\\frawk\\target\\debug\\build\\tikv-jemalloc-sys-1f5ef3a81eb3f627\\out"
  checking for xsltproc... /usr/bin/xsltproc
  checking for x86_64-pc-win32-gcc... C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.12.25827\bin\HostX64\x64\cl.exe
  checking whether the C compiler works... no
  running: "tail" "-n" "100" "<snip>\\frawk\\target\\debug\\build\\tikv-jemalloc-sys-1f5ef3a81eb3f627\\out\\build\\config.log"
  exe=''
  exec_prefix='<snip>\frawk\target\debug\build\tikv-jemalloc-sys-1f5ef3a81eb3f627\out'
  EXEEXT=''
  EXTRA_CFLAGS=''
  EXTRA_CXXFLAGS=''
  EXTRA_LDFLAGS=''
  GREP=''
  HAVE_CXX14=''
  host_alias='x86_64-pc-win32'
  host_cpu=''
  host_os=''
  host_vendor=''
  host='x86_64-pc-win32'
  htmldir='${docdir}'
  importlib=''
  includedir='${prefix}/include'
  INCLUDEDIR='<snip>frawktargetdebugbuildtikv-jemalloc-sys-1f5ef3a81eb3f627out/include'
  infodir='${datarootdir}/info'
  INSTALL_DATA=''
  INSTALL_PROGRAM=''
  INSTALL_SCRIPT=''
  install_suffix=''
  je_=''
  JEMALLOC_CPREFIX=''
  JEMALLOC_PREFIX=''
  jemalloc_version_bugfix=''
  jemalloc_version_gid=''
  jemalloc_version_major=''
  jemalloc_version_minor=''
  jemalloc_version_nrev=''
  jemalloc_version=''
  LD_PRELOAD_VAR=''
  LD=''
  LDFLAGS='-nologo -MD -Z7 -Brepro -W4'
  LDTARGET=''
  libdir='${exec_prefix}/lib'
  LIBDIR='<snip>frawktargetdebugbuildtikv-jemalloc-sys-1f5ef3a81eb3f627out/lib'
  libdl=''
  libexecdir='${exec_prefix}/libexec'
  LIBOBJS=''
  libprefix=''
  LIBS=''
  link_whole_archive=''
  LM=''
  localedir='${datarootdir}/locale'
  localstatedir='${prefix}/var'
  LTLIBOBJS=''
  mandir='${datarootdir}/man'
  MANDIR='<snip>frawk\target\debug\build\tikv-jemalloc-sys-1f5ef3a81eb3f627\out/share/man'
  MKLIB=''
  NM=''
  o=''
  OBJEXT=''
  objroot=''
  oldincludedir='/usr/include'
  PACKAGE_BUGREPORT=''
  PACKAGE_NAME=''
  PACKAGE_STRING=''
  PACKAGE_TARNAME=''
  PACKAGE_URL=''
  PACKAGE_VERSION=''
  PATH_SEPARATOR=':'
  pdfdir='${docdir}'
  PIC_CFLAGS=''
  PREFIX='<snip>\frawk\target\debug\build\tikv-jemalloc-sys-1f5ef3a81eb3f627\out'
  prefix='<snip>\frawk\target\debug\build\tikv-jemalloc-sys-1f5ef3a81eb3f627\out'
  private_namespace=''
  program_transform_name='s,x,x,'
  psdir='${docdir}'
  RANLIB=''
  rev='2'
  RPATH_EXTRA=''
  RPATH=''
  sbindir='${exec_prefix}/sbin'
  sharedstatedir='${prefix}/com'
  SHELL='/bin/sh'
  so=''
  SOREV=''
  SPECIFIED_CFLAGS='-nologo -MD -Z7 -Brepro -W4'
  SPECIFIED_CXXFLAGS=''
  srcroot=''
  sysconfdir='${prefix}/etc'
  target_alias=''
  TEST_LD_MODE=''
  XSLROOT=''
  XSLTPROC='/usr/bin/xsltproc'

  ## ----------- ##
  ## confdefs.h. ##
  ## ----------- ##

  /* confdefs.h */
  #define PACKAGE_NAME ""
  #define PACKAGE_TARNAME ""
  #define PACKAGE_VERSION ""
  #define PACKAGE_STRING ""
  #define PACKAGE_BUGREPORT ""
  #define PACKAGE_URL ""

  configure: exit 77

  --- stderr
  configure: error: in `<snip>/frawk/target/debug/build/tikv-jemalloc-sys-1f5ef3a81eb3f627/out/build':
  configure: error: C compiler cannot create executables
  See `config.log' for more details
  thread 'main' panicked at 'command did not execute successfully: "sh" "<snip>/frawk/target/debug/build/tikv-jemalloc-sys-1f5ef3a81eb3f627/out/build/configure" "--disable-cxx" "--with-jemalloc-prefix=_rjem_" "--with-private-namespace=_rjem_" "--host=x86_64-pc-win32" "--build=x86_64-pc-win32" "--prefix=C:\\Users\\K0RYU\\workspace\\frawk\\target\\debug\\build\\tikv-jemalloc-sys-1f5ef3a81eb3f627\\out"
  expected success, got: exit code: 77', C:\Users\K0RYU\.cargo\registry\src\github.com-1ecc6299db9ec823\tikv-jemalloc-sys-0.4.2+5.2.1-patched.2\build.rs:333:9
  note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
warning: build failed, waiting for other jobs to finish...
error: build failed

Basically, jemalloc isn't something we should expect to compile on Windows right now. I think this is unfortunate, given that it's currently impossible to define a feature that only is default for a specific platform.

The Overview doc says "I suspect [frawk] would run much slower on a 64-bit non-x86 architecture." Does that mean it frawk would be slower on ARM than frawk on x84, or that frawk would be slower than gawk/mawk/etc? In other words, would frawk still be faster, or should it be avoided on ARM if performance is a goal?

Scientific notation is not supported in awk script, but works for parsing files with scientific notation.

$ printf '6.18163e-27\n1.80782e-40\n2.38296e-05\n1.92843e-09\n7.37465e-39\n' | frawk '$1 <= 1e-8'

$ printf '6.18163e-27\n1.80782e-40\n2.38296e-05\n1.92843e-09\n7.37465e-39\n' | frawk '$1 <= 1E-8'


$ printf '6.18163e-27\n1.80782e-40\n2.38296e-05\n1.92843e-09\n7.37465e-39\n' | frawk '$1 <= 10^-8'
6.18163e-27
1.80782e-40
1.92843e-09
7.37465e-39

# It understands scientific notation when parsing files itself.
$ printf '6.18163e-27\n1.80782e-40\n2.38296e-05\n1.92843e-09\n7.37465e-39\n' | frawk '{ if ($1 <= 10^-8) { print $1, ($1 + 0.0); } }'
6.18163e-27 6.18163e-27
1.80782e-40 1.80782e-40
1.92843e-09 1.92843e-9
7.37465e-39 7.37465e-39


$ printf '6.18163e-27\n1.80782e-40\n2.38296e-05\n1.92843e-09\n7.37465e-39\n' | gawk '$1 <= 1e-8'
6.18163e-27
1.80782e-40
1.92843e-09
7.37465e-39

$ printf '6.18163e-27\n1.80782e-40\n2.38296e-05\n1.92843e-09\n7.37465e-39\n' | gawk '$1 <= 1E-8'
6.18163e-27
1.80782e-40
1.92843e-09
7.37465e-39

$ printf '6.18163e-27\n1.80782e-40\n2.38296e-05\n1.92843e-09\n7.37465e-39\n' | gawk '$1 <= 10^-8'
6.18163e-27
1.80782e-40
1.92843e-09
7.37465e-39